Early metabolic risk in COPD

Transcription

Early
metabolic risk
in COPD
Bram van den Borst
EarlymetabolicriskinCOPD
Copyright©2013BramvandenBorst,Maastricht
Coverdesign TijsvandenBorst
Layout
BramvandenBorst
Production Proefschriftmaken.nl Publisher
UitgeverijBOXPress ISBN 978‐90‐8891‐604‐5 The research described in this thesis was performed at the Department of
RespiratoryMedicine,NUTRIMSchoolforNutrition,Toxicology,andMetabolism,
at Maastricht University Medical Center+, Maastricht, The Netherlands; and at
the Laboratory of Epidemiology, Demography and Biometry, National Institute
onAging,Bethesda,MD,USA.
The work described in this thesis was performed within the framework of the
Dutch Top Institute Pharma, project ‘Transition of systemic inflammation into
multiorganpathology,T1‐201.’PartnersinthisprojectareMaastrichtUniversity
Medical Center+, University Medical Center Groningen, University Medical
CenterUtrecht,GlaxoSmithKline,AstraZenecaNederlandB.V.,DanoneResearch,
andNycomed(nowTakeda).
ThepublicationofthisthesiswasfinanciallysupportedbyGlaxoSmithKlineB.V.,
Nutricia Advanced Medical Nutrition, Danone Research – Centre for
SpecialisedNutrition, and Chiesi Pharmaceuticals B.V. EarlymetabolicriskinCOPD
Proefschrift
Terverkrijgingvandegraadvandoctor
aandeUniversiteitMaastricht,
opgezagvandeRectorMagnificus,Prof.dr.L.L.G.Soete,
volgenshetbesluitvanhetCollegevanDecanen,
inhetopenbaarteverdedigenop
vrijdag17mei2013om12:00uur
door
BramvandenBorst
Geborenop18juli1983teEindhoven
Promotor
Prof.dr.ir.A.M.W.J.Schols
Co‐promotor
Dr.H.R.Gosker
Beoordelingscommissie
Prof.dr.M.K.Hesselink(voorzitter)
Prof.dr.M.Laville(UniversitédeLyon,France)
Prof.dr.D.S.Postma(UniversitairMedischCentrumGroningen)
Prof.dr.C.D.A.Stehouwer
Prof.dr.E.E.Blaak
CONTENTS
Chapter1 Generalintroduction
7
Publishedinpartas:Chronicobstructivepulmonarydisease;co‐morbiditiesand
systemicconsequences.BookChapter.2011HumanaPress.Pp171‐92
Chapter2 Is age‐related decline in lean mass and physical function
acceleratedbyobstructivelungdiseaseorsmoking?
Thorax2011;66(11):961‐9
17
Chapter3 Lossofquadricepsmuscleoxidativephenotypeanddecreased
enduranceinpatientswithmild‐to‐moderateCOPD
JournalofAppliedPhysiology2012Jul19[Epubaheadofprint]
41
Chapter4 Low‐gradeadiposetissueinflammationinpatientswithmild‐
to‐moderateCOPD
AmericanJournalofClinicalNutrition2011;94(6):1504‐12
63
Chapter5 The influence of abdominal visceral fat on inflammatory
pathwaysandmortalityriskinobstructivelungdisease
89
Chapter6 Characterizationoftheinflammatoryandmetabolicprofileof 109
adiposetissueinamousemodelofchronichypoxia
JournalofAppliedPhysiology2013Mar28[Epubaheadofprint]
Chapter7 Dietary fiber and fatty acids in COPD risk and progression: 129
asystematicreview
Submitted
Chapter8 Discussion
Centralfatandperipheralmuscle:partnersincrimeinCOPD
AmericanJournalofRespiratoryandCriticalCareMedicine2013;187(1):8‐13
145
Summary
157
Samenvatting
161
Dankwoord
167
CurriculumVitae
171
Listofpublications
175
CHAPTER1
Generalintroduction
BramvandenBorst,HarryR.Gosker,AnnemieM.W.J.Schols
Inpartpublishedas:
Body composition abnormalities in COPD. Book Chapter in: Chronic
obstructive pulmonary disease; co‐morbidities and systemic
consequences.2011HumanaPress.Pp171‐92
7
GENERALINTRODUCTION
AnintroductiontoCOPD
Chronic obstructive pulmonary disease (COPD) is a common disease and is
associated with high burden. Symptoms include shortness of breath, exercise
limitation and cough. Sixty‐four million people suffer from COPD worldwide,
being the fourth leading cause of mortality [1]. In the Netherlands, more than
320.000peoplesufferfromCOPD,causing4%ofalldeaths[2].
Cigarette smoke is the primary cause of COPD, but other risk factors have
beenidentifiedincludingoccupationalexposuretotoxicgases,airpollutionand
genetic predisposition [3]. COPD patients display an enhanced and persistent
inflammatory response to inhaled particles which is believed to underlie local
pulmonarypathology[4].Chronicbronchitisandemphysemaarethetwomajor
pathological features of COPD. Briefly, chronic bronchitis is characterized by
thickening of the airway walls and increased mucus production, whereas
emphysema denotes the destruction of alveolar walls. These disruptions of the
pulmonary architecture ultimately lead to the obstruction of airflow during
exhalation, which is usually progressive in time. As the degree of airflow
limitation can be measured relatively easily and with good reproducibility
throughspirometry,itiscommonlyusedasthesinglemeasureofCOPDseverity.
However,clinicalobservationsandadvancingresearchhaveclearlyshownthat
thedegreeofairflowlimitationispoorlycorrelatedwithactualdiseaseburden.
Clinicalheterogeneityandbodycomposition
Clinical heterogeneity is an important characteristic of COPD that complicates
treatment. Different body composition phenotypes contribute to this
heterogeneity and reflect important extrapulmonary disease features in COPD,
asillustratedbythefollowingcases.
Case1:thecachecticCOPDpatient
Patient A, 63 years old, has a forced expiratory volume in 1 sec (FEV1) and
diffusion capacity for carbon monoxide (DLCO) of 52% and 37% of predicted,
respectively.Sincediagnosisshehaslost18kgofbodymassandnowsheweighs
only 40 kg resulting in a current body mass index (BMI) of 15.2 kg/m2. Body
composition measures reveal a fat‐free mass index (FFMI) and fat mass index
(FMI)of12.1and3.1kg/m2,respectively.BoththeFFMIandFMIarebelowthe
5th percentile of reference values, indicative of severe depletion. After having
smoked60packyearssheremainsunmotivatedtoquit.Lastyear,shehadbeen
admitted to the hospital twice for an acute exacerbation and has also been
diagnosedwithosteoporosis.Duetoherweaknesssheishomeboundandspends
mostofthedaysedentary.Shehasanarterialoxygentensionof6.9kPa,witha
8
CHAPTER1
carbondioxidepressureof5.6kPa,consistentwitharterialhypoxaemiawithout
respiratoryinsufficiency.
Case2:thesarcopenicCOPDpatient
AlthoughpatientB,72yearsold,wasdiagnosedwithCOPDmanyyearsago,he
remainsasmokerwith45packyears.TheFEV1hasbeenstablearound49%of
predicted.HisBMIhasremainedstableforyearsat27kg/m2,yethisweighthas
recently started to decrease slowly. The FFMI and FMI measure 17.8 (10‐25th
percentile) and 9.2 (90‐95th percentile) kg/m2, respectively. His daily 20‐min
walkbecomesincreasinglydifficultduetoafeelingofwholebodyweakness.He
questionswhetherthisisduetoagingorifthisiscausedbytheCOPD.
Case3:theobeseCOPDpatient
PatientCis58yearsoldandwasdiagnosedwithCOPDfouryearsago.Shehasa
part‐time job and although she experiences mild shortness of breath upon
exertion,sheisanactivememberofawalkingclub.TheFEV1andDLCOmeasure
56%and81%ofpredicted,respectively.HerBMIis33kg/m2whichisclassified
as obese. FFMI and FMI measure 18.2 (75‐90th percentile) and 12.7 (90‐95th
percentile) kg/m2, respectively. Her weight has been stable for decades. Peak
oxygen uptake during an incremental cycling test is 89% of predicted during
which no arterial desaturation is observed. The COPD is controlled by her
general practitioner and after having smoked 45 pack years she succeeded
quitting,afterwhichshegainedsomemoreweight.Shehasalsobeendiagnosed
withhypertension.
Notwithstanding that the lung is the major organ of pathology in COPD, these
casesillustratethreepatientswithcomparabledegreesofairflowlimitationbut
with widely varying clinical presentations. Skeletal muscle wasting and
weakness, osteoporosis, but also obesity are common in COPD patients, and
differentially contribute to disease burden throughout the disease course. Only
veryrecently,however,ithasbeenofficiallyrecognizedinthedefinitionofCOPD
that “…comorbidities contribute to theoverall severity in individual patients” [4].
Infact,ithasbeenshownthatCOPDpatientshaveonaveragetwotothreeother
chronic conditions, of which skeletal muscle dysfunction and cardiovascular
comorbiditiesareparticularlycommon[5].
SkeletalmuscledysfunctioninCOPD
Skeletal muscle dysfunction is probably the most investigated extrapulmonary
manifestation in COPD. Muscle wasting, along with general weight loss, is
particularly outspoken in but not restricted to advanced COPD, and has been
9
GENERALINTRODUCTION
identified as a poor prognostic factor independently of the degree of airflow
limitation[6].Importantly,weightgaincanreversethisprognosis[7].
With aging, particularly in the eighth decade of life, a change in body
composition occurs which has been termed sarcopenia [8]. Sarcopenia is the
processofadecreasingmusclemassduringwhichfatmassincreasesorremains
stable, and during which body weight can remain stable. This hidden loss of
musclemasshasbeenidentifiedinasignificantproportionofCOPDpatientsin
cross‐sectionalstudies(upto25%[9‐11]),butwhethersarcopeniaoccursinan
acceleratedfashioninCOPDneedstobeinvestigatedinlongitudinalstudieswith
appropriatecontrolgroups.
Inadditiontomusclewasting,intrinsicalterationsintheskeletalmuscleof
COPDpatientshavebeenidentified.Thesearecharacterizedbyashiftfromless
oxidative towards more glycolytic fibers, decreased oxidative enzyme capacity
andmitochondrialdysfunctionleadingtoearlyfatiguedmusclesanddecreased
exercise capacity [12, 13]. Although the intrinsic alterations underlying muscle
dysfunction have been well‐documented in patients with advanced COPD, few
studieshaveexploredthemechanismsofskeletalmuscledysfunctioninearlier
stagesofCOPD.
CardiovascularcomorbidityinCOPD
Throughout all severity stages COPD patients have an approximate 2 to 3‐fold
increased risk of cardiovascular disease (CVD), which goes beyond cigarette
smokeexposure[14].VariousfactorshavebeensuggestedtoincreaseCVDrisk
in COPD patients, including generic risk factors (unhealthy lifestyle) as well as
COPD‐related factors such as low‐grade systemic inflammation, hypoxaemia,
oxidativestress,andsystemicelastinabnormalities.TheobservationthatCVDis
even the leading cause of mortality in COPD patients with mild‐to‐moderate
airflowlimitation[15],stressestheneedtoidentifyunderlyingmechanismsthat
drivecardiovascularriskparticularlyinthissubgroupofCOPDpatients.
ThemetabolicsyndromeanddiabetesappeartobemorecommoninCOPD
patientscomparedtocontrolpopulations,withmild‐to‐moderateCOPDpatients
being more affected than severe COPD patients [16]. Likely, these co‐morbid
conditionsalsocontributetoCVDrisk.Peripheralskeletalmusclemitochondrial
dysfunctionhasbeenassociatedwithinsulinresistanceindiabetesandobesity
[17], and excessive abdominal visceral fat accumulation has been suggested to
play an upstream role in low‐grade systemic inflammation, hypertension and
atheroscleroticdiseaseinepidemiologicalpopulationstudies[18].Itisunknown
whether these metabolic alterations already exist in patients with mild‐to‐
moderateCOPD.
10
CHAPTER1
Low‐gradesystemicinflammationinCOPD:aroleforadipose
tissue?
Pathophysiological mechanisms linking COPD and extrapulmonary
manifestations remain incompletely understood. As several studies have
suggestedthatsmokingdoesnotfullyaccountfortheseassociations,itappears
thatotherfactorsareimplicated[19].Ithasbeensuggestedthatincreasedlevels
of circulating inflammatory mediators, also referred to as low‐grade systemic
inflammation, are implicated in the risk of extrapulmonary manifestations [5].
Indeed, many studies have reported systemic inflammation in COPD [20], and
inflammatory cytokines have been proposed to play a pivotal role in the
development of e.g. skeletal muscle dysfunction [21, 22], atherosclerosis [23]
andosteoporosis[24].Knowingthesourceoflow‐gradesystemicinflammation
isimportantasitmayguidetargetedfuturetherapeuticstrategies.Classicallyit
is thought that enhanced levels of plasma cytokines in COPD originate from
leakage from the pulmonary compartment [5]. However, to date no convincing
evidence for this in stable disease has been published, suggesting that other
sourcesofinflammationexist.
Whereas the adipose organ was always considered a passive storage
compartment for energy, it is now widely recognized as a highly potent
production site of inflammation. Adipose tissue produces and secretes
inflammatorymediators,whichhavebeentermedadipokinesoradipocytokines.
Interleukin‐6, TNF‐α, leptin and adiponectin are some examples [25].
Adipocytokines act both locally (paracrine) as well as systemically (endocrine)
andinfluenceinflammationandmetabolism.Obesity,alsointheabsenceofany
otherchroniccondition,isassociatedwithlow‐gradesystemicinflammation.In
thefieldofobesityresearch,majoradvanceshavebeenmadeinunderstanding
the relation between excessive fat mass and low‐grade systemic inflammation.
Adiposetissuehypoxiaresultingfromhypertrophyingadipocytesandimpaired
neovascularisationhasbeensuggestedtotriggeradiposetissueinflammationin
obesity. Therefore, it can be speculated that adipose tissue hypoxia, as a
consequence of hypoxaemia (chronic or intermittent resulting from
desaturations), also triggers adipose tissue inflammation in COPD. Moreover,
overweight and obesity are common in COPD [26], particularly in those with
mild‐to‐moderate airflow limitation [27]. Although several recent studies have
suggestedapositiverelationbetweenBMIandlow‐gradesystemicinflammation
[28, 29], the relations between (changes in) fat mass, body fat distribution,
adipose tissue inflammation, and systemic inflammation remain poorly
understoodinCOPD.
11
GENERALINTRODUCTION
Lifestylefactors
Exposure to cigarette smoke is not only the major risk factor for COPD
development;ithasalsobeenassociatedwithlow‐gradesystemicinflammation
intheabsenceofCOPD,suggestingthatsmokingaccountsforatleastpartofthe
systemic manifestations in COPD. Also, smokers in the general population are
usuallyleaner[30]andscoreworseonphysicalperformancetestscomparedto
never‐smokers [31]. It remains unclear, however, to what extent smoking
accountsforthechangesinbodycompositionandphysicalfunctioninginCOPD.
COPD patients experience major difficulties in performing physical activity.
Dyspnea upon exertion partially contributes to physical inactivity. Being
physicallyinactiveisknowntohavedeconditioningeffectsonperipheralskeletal
muscleswhich,inturn,furtheracceleratethisdownwardspiral.Althoughithas
beenwell‐describedthatCOPDpatientsperformlesstotaldailyphysicalactivity,
the quality of the activity has been scarcely investigated. In fact, regular
moderate‐to‐vigorous physical exercise lasting at least 30 minutes per day has
beenrecommendedinordertogainhealthbenefits[32].Itisunknownhowthe
quality of physical activity in COPD relates to these recommendations, and to
whatextentthisisassociatedwithmusculardeconditioningoradversebodyfat
distribution.
Western society is characterized by an ‘obesogenic’ environment in which
poor dietary quality contributes to low health status. In fact, the European
Respiratory Society has recently called for more research on whether diet
influences the severity of respiratory diseases [33]. A ‘Western diet’ rich in
refined grains, red meat, desserts and French fries has been associated with
increased risk of COPD development, independent of smoking [34], and,
conversely, a dietary pattern rich in fruit, vegetables, oily fish and wholemeal
cerealsmayprotectagainstimpairedlungfunctionandCOPD[35].Toimprove
nutritional recommendations in COPD, there is a need to examine the specific
relations between dietary components and outcomes in COPD. In this context,
dietary fiber and fatty acids are of particular interest given their potential to
influenceinflammatoryprocesses.
12
CHAPTER1
Aimofthisthesis
The general aim of this thesis is to study the reciprocal relation between body
composition and low‐grade systemic inflammation in mild‐to‐moderate COPD,
andtoexplorethepotentialinvolvementofanunhealthylifestyleinthisrelation
(Figure1).
Figure1.Schematicrepresentationofthegeneralaimofthisthesis.
Theresearchquestionsthatareaddressedinthisthesisare:
1. Are changes in body composition in COPD different from the aging‐related
naturalcourseofachangingbodycomposition?
2. Aremild‐to‐moderateCOPDpatientsatanearlymetabolicriskasreflectedin:
a. impairedskeletalmusclephenotypeandmetabolism;
b. adipose tissue inflammation contributing to low‐grade systemic
inflammation;
c. excessiveabdominalvisceralfatmass?
3. What is the influence of smoking, diet and physical activity on early
metabolicriskinCOPDandonCOPDoutcomes?
Outlineofthisthesis
Chapter2presentsacombinedcross‐sectionaland7‐yearlongitudinalanalysis
ofbodycomposition,musclestrengthandphysicalfunctioningofCOPDpatients
and control persons with normal lung function stratified by smoking status
(researchquestions1and3).
Chapter3presentsacross‐sectionalcomparisonofquadricepsmuscleoxidative
phenotype,quadricepsfatigueanddailyphysicalactivitylevelbetweenpatients
withmild‐to‐moderateCOPDandmatchedhealthysubjects(researchquestions
2aand3).
13
GENERALINTRODUCTION
Chapter 4 presents a cross‐sectional analysis of abdominal subcutaneous
adipose tissue inflammatory status and its relation with circulating
adipocytokine levels in patients with mild‐to‐moderate COPD and matched
healthysubjects(researchquestion2b).
Chapter 5 presents a comprehensive analysis of differences in abdominal fat
distribution between COPD patients and matched controls, and assesses the
associations between visceral fat mass and systemic inflammation and their
predictive value for 9‐year mortality. This chapter also investigates differences
in dietary composition and physical activity between these COPD patients and
matchedcontrols(researchquestions2b,2cand3).
Chapter 6 presents an experimental study in mice in which the inflammatory
and metabolic changes in abdominal subcutaneous and visceral adipose tissue
arecharacterizedafterexposuretochronichypoxia(researchquestion2b).
Chapter7presentsasystematicliteraturereviewontheassociationsbetween
dietaryintakeoffiberandspecificfattyacidsandCOPDetiologyandprogression
(researchquestion3).
In Chapter 8 the metabolic abnormalities and adverse lifestyle components in
COPDpatientsidentifiedinthisthesisarediscussed.
14
CHAPTER1
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
World Health Organization. The global burden of disease: 2004 update. 2008; Available from
http://www.who.int/healthinfo/global_burden_disease/GBD_report_2004update_full.pdf.
Boezen HM, Postma DS, Smit HA. COPD samengevat. In: Volksgezondheid Toekomst Verkenning, Nationaal
KompasVolksgezondheid.Bilthoven:RIVM.2008.
SorianoJB,Rodriguez‐RoisinR.Chronicobstructivepulmonarydiseaseoverview:epidemiology,riskfactors,
andclinicalpresentation.ProcAmThoracSoc.2011;8(4):363‐7.
Global Initiative for Chronic Obstructive Lung Disease. Global strategy for the diagnosis, management, and
prevention of chronic obstructive pulmonary disease. Revised 2011. Available from
http://www.goldcopd.org/uploads/users/files/GOLD_Report_2011_Feb21.pdf.
BarnesPJ,CelliBR.SystemicmanifestationsandcomorbiditiesofCOPD.EurRespirJ.2009;33(5):1165‐85.
ScholsAM,WoutersEF.Nutritionalconsiderationsinthetreatmentofchronicobstructivepulmonarydisease.
ClinNutr.1995;14(2):64‐73.
Schols AM, Slangen J, Volovics L, Wouters EF. Weight loss is a reversible factor in the prognosis of chronic
obstructivepulmonarydisease.AmJRespirCritCareMed.1998;157(6Pt1):1791‐7.
Fielding RA, Vellas B, Evans WJ, Bhasin S, Morley JE, Newman AB, Abellan van Kan G, Andrieu S, Bauer J,
Breuille D, et al. Sarcopenia: an undiagnosed condition in older adults. Current consensus definition:
prevalence, etiology, and consequences. International working group on sarcopenia. J Am Med Dir Assoc.
2011;12(4):249‐56.
ScholsAM,SoetersPB,DingemansAM,MostertR,FrantzenPJ,WoutersEF.Prevalenceandcharacteristicsof
nutritionaldepletion in patientswith stable COPD eligiblefor pulmonary rehabilitation. Am Rev Respir Dis.
1993;147(5):1151‐6.
Schols AM, Broekhuizen R, Weling‐Scheepers CA, Wouters EF. Body composition and mortality in chronic
obstructivepulmonarydisease.AmJClinNutr.2005;82(1):53‐9.
VestboJ,PrescottE,AlmdalT,DahlM,NordestgaardBG,AndersenT,SorensenTI,LangeP.Bodymass,fat‐free
bodymass,andprognosisinpatientswithchronicobstructivepulmonarydiseasefromarandompopulation
sample:findingsfromtheCopenhagenCityHeartStudy.AmJRespirCritCareMed.2006;173(1):79‐83.
AllaireJ,MaltaisF,DoyonJF,NoelM,LeBlancP,CarrierG,SimardC,JobinJ.Peripheralmuscleenduranceand
theoxidativeprofileofthequadricepsinpatientswithCOPD.Thorax.2004;59(8):673‐8.
GoskerHR,ZeegersMP,WoutersEF,ScholsAM.Musclefibretypeshiftinginthevastuslateralisofpatients
with COPD is associated with disease severity: a systematic review and meta‐analysis. Thorax.
2007;62(11):944‐9.
Sin DD, Man SF. Why are patients with chronic obstructive pulmonary disease at increased risk of
cardiovascular diseases? The potential role of systemic inflammation in chronic obstructive pulmonary
disease.Circulation.2003;107(11):1514‐9.
McGarveyLP,JohnM,AndersonJA,ZvarichM,WiseRA.Ascertainmentofcause‐specificmortalityinCOPD:
operationsoftheTORCHClinicalEndpointCommittee.Thorax.2007;62(5):411‐5.
Watz H, Waschki B, Kirsten A, Muller KC, Kretschmar G, Meyer T, Holz O, Magnussen H. The metabolic
syndromeinpatientswithchronicbronchitisandCOPD:frequencyandassociatedconsequencesforsystemic
inflammationandphysicalinactivity.Chest.2009;136(4):1039‐46.
Kelley DE, He J, Menshikova EV, Ritov VB. Dysfunction of mitochondria in human skeletal muscle in type 2
diabetes.Diabetes.2002;51(10):2944‐50.
KishidaK,FunahashiT,MatsuzawaY,ShimomuraI.Visceraladiposityasatargetforthemanagementofthe
metabolicsyndrome.AnnMed.2012;44(3):233‐41.
PatelAR,HurstJR.Extrapulmonarycomorbiditiesinchronicobstructivepulmonarydisease:stateoftheart.
ExpertRevRespirMed.2011;5(5):647‐62.
Gan WQ, Man SF, Senthilselvan A, Sin DD. Association between chronic obstructive pulmonary disease and
systemicinflammation:asystematicreviewandameta‐analysis.Thorax.2004;59(7):574‐80.
RemelsAH,GoskerHR,SchrauwenP,HommelbergPP,SliwinskiP,PolkeyM,GaldizJ,WoutersEF,LangenRC,
ScholsAM.TNF‐alphaimpairsregulationofmuscleoxidativephenotype:implicationsforcachexia?FASEBJ.
2010;24(12):5052‐62.
LangenRC,ScholsAM,KeldersMC,vanderVeldenJL,WoutersEF,Janssen‐HeiningerYM.Musclewastingand
impaired muscle regeneration in a murine model of chronic pulmonary inflammation. Am J Respir Cell Mol
Biol.2006;35(6):689‐96.
PaolettiR,GottoAM,Jr.,HajjarDP.Inflammationinatherosclerosisandimplicationsfortherapy.Circulation.
2004;109(23Suppl1):III20‐6.
McLeanRR.Proinflammatorycytokinesandosteoporosis.CurrOsteoporosRep.2009;7(4):134‐9.
Trayhurn P, Wood IS. Adipokines: inflammation and the pleiotropic role of white adipose tissue. Br J Nutr.
2004;92(3):347‐55.
Vozoris N, O'Donnell DE. Prevalence, risk factors, activity limitation and health care utilization of an obese,
population‐basedsamplewithchronicobstructivepulmonarydisease.CanRespirJ.2012;19(3):e18‐24.
Agusti A, Calverley PM, Celli B, Coxson HO, Edwards LD, Lomas DA, MacNee W, Miller BE, Rennard S,
SilvermanEK,etal.CharacterisationofCOPDheterogeneityintheECLIPSEcohort.RespirRes.2010;11:122.
Eagan TM, Ueland T, Wagner PD, Hardie JA, Mollnes TE, Damas JK, Aukrust P, Bakke PS. Systemic
inflammatorymarkersinCOPD:resultsfromtheBergenCOPDCohortStudy.EurRespirJ.2010;35(3):540‐8.
15
GENERALINTRODUCTION
29. Agusti A, Edwards LD, Rennard SI, Macnee W, Tal‐Singer R, Miller BE, Vestbo J, Lomas DA, Calverley PM,
Wouters E, et al. Persistent Systemic Inflammation is Associated with Poor Clinical Outcomes in COPD: A
NovelPhenotype.PLoSOne.2012;7(5):e37483.
30. AkbartabartooriM,LeanME,HankeyCR.Relationshipsbetweencigarettesmoking,bodysizeandbodyshape.
IntJObes(Lond).2005;29(2):236‐43.
31. Strand BH, Mishra G, Kuh D, Guralnik JM, Patel KV. Smoking history and physical performance in midlife:
resultsfromtheBritish1946birthcohort.JGerontolABiolSciMedSci.2011;66(1):142‐9.
32. PateRR,PrattM,BlairSN,HaskellWL,MaceraCA,BouchardC,BuchnerD,EttingerW,HeathGW,KingAC,et
al.Physicalactivityandpublichealth.ArecommendationfromtheCentersforDiseaseControlandPrevention
andtheAmericanCollegeofSportsMedicine.JAMA.1995;273(5):402‐7.
33. European Respiratory roadmap‐ recommendations for the future of respiratory medicine. 2012; Available
from:http://www.ersroadmap.org.
34. Varraso R, Fung TT, Hu FB, Willett W, Camargo CA. Prospective study of dietary patterns and chronic
obstructivepulmonarydiseaseamongUSmen.Thorax.2007;62(9):786‐91.
35. ShaheenSO,JamesonKA,SyddallHE,SayerAA,DennisonEM,CooperC,RobinsonSM.Therelationofdietary
patternswithadultlungfunctionandCOPD.EurRespirJ.2010;36(2):277‐84.
16
CHAPTER1
CHAPTER2
Is age‐related decline in lean mass
and physical function accelerated by
obstructivelungdiseaseorsmoking?
BramvandenBorst,AnnemarieKoster,BinbingYu,HarryR.Gosker,
Bernd Meibohm, DouglasC. Bauer, StephenB. Kritchevsky,Yongmei
Liu,AnneB.Newman,TamaraB.Harris,AnnemieM.W.J.Schols
FortheHealth,AgingandBodyComposition(HealthABC)Study
Thorax2011;66(11):961‐9
17
SMOKING,SARCOPENIAANDCOPD
Abstract
Background and aims: Cross‐sectional studies suggest that obstructive lung
disease (OLD) and smoking affect lean mass and mobility. We aimed to
investigate whether OLD and smoking accelerate aging‐related decline in lean
massandphysicalfunctioning.
Methods: 260 persons with OLD (FEV1 63±18 % of predicted), 157 smoking
controls (FEV1 95±16 % of predicted), 866 formerly smoking controls (FEV1
100±16 % of predicted) and 891 never‐smoking controls (FEV1 104±17 % of
predicted)participatingintheHealth,AgingandBodyComposition(ABC)Study
werestudied.Atbaseline,themeanagewas74±3yandparticipantsreportedno
functional limitations. Baseline and seven‐year longitudinal data were
investigatedofbodycomposition(bydual‐energyX‐rayabsorptiometry),muscle
strength (by hand and leg dynamometry) and short physical performance
battery(SPPB).
Results: Compared to never‐smoking controls, OLD persons and smoking
controlshadasignificantlylowerweight,fatmass,leanmassandbonemineral
content(BMC)atbaseline(P<.05).Whilethelossofweight,fatmass,leanmass
and strength was comparable between OLD persons and never‐smoking
controls,theSPPBdeclined0.12points/yrfasterinOLDmen(P=.01)andBMC4
g/yrfasterinOLDwomen(P=.02).Insmokingcontrols,onlyleanmassdeclined
0.1 kg/yr faster in women (P=.03) and BMC 8 g/yr faster in men (P=.02)
comparedtonever‐smokingcontrols.
Conclusions:Initiallywell‐functioningolderadultswithmild‐to‐moderateOLD
and smokers without OLD have a comparable compromised baseline profile of
body composition and physical functioning, while seven‐year longitudinal
trajectoriesaretoalargeextentcomparabletothoseobservedinnever‐smokers
withoutOLD.Thissuggestsacommoninsultearlierinliferelatedtosmoking.
18
CHAPTER2
Introduction
Chronic obstructive pulmonary disease (COPD) affects approximately 14% of
olderadultsandisrelatedtoincreasedimmobility,hospitalizationsandresulting
healthcarecosts[1].COPDischaracterizedbyirreversibleairflowlimitationand
is associated with an abnormal inflammatory response of the lungs mainly to
cigarette smoke [2]. It has been increasingly recognized as a disease with
marked extrapulmonary manifestations such as muscle wasting and weakness,
osteoporosis and cardiovascular disease [3], in which systemic inflammation is
thought to play an important role [4, 5]. It remains unclear, however, if these
systemicmanifestationsareco‐morbidconditionsrelatedtoacommonnoxious
exposureorifthesearesystemicconsequencesofthediseaseitself.
So far skeletal muscle weakness and wasting are the most studied and
established extra‐pulmonary determinants of impaired mobility and increased
mortalityinCOPD[6].Putativepathophysiologicalmechanismsofalteredbody
composition and physical functioning in COPD have originated from cross‐
sectionalstudies[7‐10]andfromlongitudinalcasestudieswithoutappropriate
control groups [11‐14]. However, since body composition changes and
dysregulation of inflammation are also common with aging [15, 16] and since
studied COPD populations often consist of older adults (60+), inclusion of
matched non‐COPD control groups is crucial to unravel the effect of airflow
obstructiononthepatternandprogressionofchangesinbodycompositionand
physical function, as well as the potential modulating role of systemic
inflammatory markers. To our best knowledge, such studies are currently
lacking. Cross‐sectional studies in healthy smokers have suggested detrimental
effectsofsmokingonskeletalmusclefunction[17,18].Increasedsystemiclevels
ofpro‐inflammatorymarkerswereobservedinhealthysmokers[19],butalsoin
formersmokers[20].Collectively,thesedatasuggestthatalsointheabsenceof
chronic airflow limitation, smoking may enhance skeletal muscle wasting and
accelerate the decline in physical functioning, and that systemic inflammation
couldbeamodulator.
Wehypothesizedthatolderadultswithobstructivelungdisease(OLD)and
smokers with normal lung function have accelerated loss of lean mass and
physical functioning compared to a never‐smoking population. Systemic
inflammatory markers could contribute to the association with body
composition and physical functioning. We therefore examined seven‐year
longitudinal changes in body composition and physical functioning in older
personswithOLDandinanon‐OLDpopulationstratifiedbysmokingstatus.
19
Methods
Studypopulation
ThisstudywasperformedinpersonsparticipatingintheHealth,AgingandBody
Composition (Health ABC) study. This is a longitudinal study of 3075 well‐
functioning black and white men and women between the ages of 70‐79 years
residinginandnearPittsburgh,PennsylvaniaandMemphis,Tennessee.Baseline
data was obtained in 1997/1998 through in‐person interview and clinic based
examination. Inclusion criteria were no reported difficulty walking a quarter
mile,climbing10stepswithoutresting,orperformingmobility‐relatedactivities
of daily living. Exclusion criteria were any life‐threatening condition,
participation in any research study involving medications or modification of
eatingorexercisehabits,planstomovefromthegeographicareawithin3years,
and difficulty in communicating with the study personnel or cognitive
impairment. The Health ABC study protocol was approved by the Institutional
Review Boards of the clinical sites. Written informed consent was obtained for
allparticipants.ThecurrentstudypresentsananalysisoftheHealthABCdataset
using 7‐year follow‐up data from participants who met the criteria for
obstructive lung disease (OLD) (n=260) and control participants with normal
lungfunction(n=1914).Participantswithmissingbodycompositionmeasuresat
baseline and those with no measures of systemic inflammatory markers were
excluded(Figure1).
Measures
Lungfunction,smokingstatusandsmokinghistory
Lung function was assessed according to international standards as previously
reported [10]. OLD was defined at baseline as reduced (i.e. < lower limit of
normal,LLN)forcedexpiratoryvolumein1s(FEV1)/forcedvitalcapacity(FVC)
as determined by age, sex, and race‐normalized values [10, 21, 22]. Cigarette
smokingstatus(smoker,formersmoker,never‐smoker)wasrecordedbasedon
self‐report. At baseline, smokers, former smokers and never‐smokers without
OLD (i.e. FEV1/FVC≥LLN) were designated as the smoking controls, formerly
smoking controls and never‐smoking controls, respectively. Participants with
restrictive lung disease (FEV1/FVC≥LLN but FVC<LLN) were excluded. At
baseline,thenumberofpackyearssmoked(1packyear=20cigarettes/dayfor
1year)wasobtainedbasedonself‐report.
Lung function was repeated in years 5 and 8. Participants with valid lung
functionmeasuresbothatbaselineandinyear5and/or8wereincludedinthe
analysis of lung function decline. Consequently, lung function decline could be
analyzed in n=1658 participants over a mean (SD) period of 6.2 (1.3) years
including2.6(0.5)validmeasurementsperparticipant.
20
CHAPTER2
Figure1.Flowchartofparticipantselectionatbaseline.
Abbreviations:ATS,AmericanThoracicSociety;FEV1,forcedexpiratoryvolumein1s;FVC,forcedvital
capacity;LLN,lowerlimitofnormal;OLD,obstructivelungdisease.
Bodycomposition
Height was measured using a wall mounted stadiometer. Body weight was
assessed to the nearest 0.1 kg using a standard balance beam scale and body
mass index (BMI) was calculated as weight/height2 (kg/m2). Weight was
measuredinyears1to6and8.WholebodydualenergyX‐rayabsorptiometry
(DXA, Hologic 4500A software version 8.21, Waltham, MA) was applied to
retrieve whole body fat mass, lean mass, leg lean mass (left and right leg were
summed)andbonemineralcontent(BMC).DXAmeasurementswereperformed
inyears1to6and8.
Musclefunctionandmobility
In years 1, 2, 4, 6 and 8, maximal isokinetic strength of the quadriceps muscle
was assessed by a Kin‐Com 125 AP Dynamometer (Chattanooga, TN, USA) at
60°/s. The right leg was tested unless there was a contraindication. The
maximum torque was taken from three reproducible and acceptable trials.
Participants with a systolic blood pressure ≥200 mmHg or a diastolic blood
pressure≥110mmHgorwhoreportedahistoryofcerebralaneurysm,cerebral
bleeding, bilateral total knee replacement, or severe bilateral knee pain were
21
excludedfromtesting(12.7%oforiginalcohort)[15].Asaconsequence,baseline
measuresforquadricepsstrengthwereavailableforn=228OLDparticipantsand
n=1730 controls. Hand grip strength was measured using a Jamar Hydraulic
HandDynamometeratbaselineandinyears2,4,6and8.Themaximumvalues
oftherightandthelefthandweresummedfromtwotrials.Inyears1,4and6,
lower extremity function was assessed using the SPPB (Short Physical
Performance Battery) [23, 24]. The battery consists of a test of gait speed,
standingbalance,andtimetorisefromachairfivetimes.Eachitemwasscored
using a five‐point scale (0 = inability to complete test, 4 = highest level of
performance)leadingtoacombined0‐12‐pointsummaryscale.
Inflammatorymarkers
At baseline, measuresofthe cytokines interleukin (IL)‐6, tumor necrosis factor
(TNF)‐αandC‐reactiveprotein(CRP)wereobtainedfromfrozenstoredplasma
orserum(seeforfurtherdetails[25]).
Covariates
At baseline, clinic site, age, race (black/white), oral steroid use, calcium
supplementation and vitamin D supplementation were based on self‐report.
Prevalent health conditions (diabetes, cardiovascular disease and depression)
wereassessedbasedonself‐reportandmedicationinventory.Physicalactivityin
theprecedingsevendayswasassessedatbaselinebyquestionnaire[26,27].
Statisticalanalyses
Differences in descriptive characteristics at baseline between OLD persons,
smokingcontrols, formerly smoking controlsand never‐smoking controls were
tested using ANOVA for continuous variables, χ2 for categorical variables and
Kruskal‐Wallistestforcontinuousvariableswithskeweddistributions.Post‐hoc
comparisons were performed applying Bonferroni correction for multiple
testing. Change in lung function could be calculated from a maximum of three
time pointsonly, hence per participantsimple linearregression was applied to
examine the yearly change (slope) in lung function. Subsequently, the mean
slopes were tested between the study groups using ANOVA. The longitudinal
data on body composition and physical functioning were analyzed using
multilevel regression models, allowing for the intercepts and slopes to vary
between the four groups. This type of analysis takes into account the intra‐
individualcorrelationbetweenrepeatedmeasurementsandallowstheinclusion
of participants with incomplete data at follow‐up. Therefore, no imputation of
missing data was applied. The primary model included group (OLD, smoking
controls, formerly smoking controls, never‐smoking controls), age, race, clinic
22
CHAPTER2
site,examinationyear(time),oralsteroiduse(yes/no),calciumsupplementation
(yes/no),vitaminDsupplementation(yes/no),cardiovasculardisease(yes/no),
diabetes(yes/no),depression(yes/no),physicalactivity,andaninteractionterm
withtimeforallofthesevariables.Thereportedslopesarethecoefficientsofthe
interactiontermgroup*timeinwhichthenever‐smokingcontrolsservedasthe
reference group. In the second model further adjustment was made for IL‐6,
TNF‐α and CRP and their interactions with time. There were no significant
interaction effects between race and clinic site with the group variable on
changes in body composition or physical functioning (P>.05). Sample size
considerationsprecludedexaminationoftheeffectofsmokingstatuswithinthe
OLD group. The current analyses were performed for men and women
separately because the gender distribution across the study groups was
significantly different (P<.001), which was most pronounced in the formerly
smokingcontrolsandnever‐smokingcontrols.
A small proportion of participants did not have measurements of body
composition and physical functioning due to missing clinic visits while being
alive. To examine the sensitivity of the results from the multilevel regression
analyses to these missing observations, we used a joint model for the primary
outcomes and missing probabilities [28]. Statistical analyses were performed
usingPASWStatistics17.0(SPSSinc.Chicago,IL).Differencesweredeclaredto
bestatisticallysignificantatP<.05.
Results
Baseline characteristics are presented in Tables 1 (men) and 2 (women). The
degreeofairflowlimitationinOLDpersonswasmild(FEV1>70%ofpredicted),
moderate (FEV1 50‐70% of predicted) and severe (FEV1<50% of predicted) in
32%, 38% and 30% of cases in men and in 36%, 41% and 23% of cases in
women, respectively. The proportion of smokers, former smokers and never‐
smokersintheOLDpersonswas63%,28%and9%inmenand46%,27%and
27%inwomen,respectively.Comparedtonever‐smokingcontrols,thedeclinein
FEV1 was lower in OLD men (‐38±6 vs. ‐57±3 mL/yr, P=.038) but not in OLD
women (‐33±5 vs. ‐40±2 mL/yr, P=.45) (Supplemental Table 1). The formerly
smokingcontrolshadabstainedfromsmokingfor25(14)years.Theprevalence
of co‐morbid conditions was not different across the groups (P>.05). Persons
withOLDandsmokingcontrolshadcomparablelevelsofphysicalactivityasthe
never‐smokingcontrols(P>.05).Inmen,OLDpersonshadincreasedCRPandIL‐
6, and smoking controls had increased IL‐6 compared with never‐smoking
controls (P<.05). In women, smoking controls had increased IL‐6 compared to
never‐smokingcontrols(P<.05).
In both sexes, persons with OLD and smoking controls showed striking
similarities regarding lower measures of body composition and physical
functioning as compared to the never‐smoking controls (Figure 2). In general,
23
Table 1. Baseline characteristics of men in the Health ABC Study according to
obstructivelungdisease(OLD)statusandsmokingstatus.
OLD(n=150)
Smoking
controls
(n=74)
Formerly
smoking
controls
(n=534)
Never‐
smoking
controls
(n=301)
A
B
C
D
73.4(2.9)
73.0(2.5)
73.7(2.9)
74.0(2.8)
.06
‐
52
35
70
65
<.001
AC,BC,BD
60
50
47
51
.013
AC,AD
72(39‐114)
.68
‐
P
Significant
post‐hoc
comparisonsa
Demographics
Age,y
Race,%white
Site,%Memphis
Physicalactivity,
kcal/kg/wkb
68(37‐105) 59(35‐125) 67(40‐114)
Co‐morbidity,steroiduse,calciumandvitaminDsupplementation
Depression,%
7
1
8
5
.22
‐
Cardiovascular
disease,%
25
24
30
23
.10
‐
Diabetes,%
13
15
16
15
.91
‐
Oralsteroiduse,%
1
1
2
0
.49
‐
Calcium
supplementation,%
5
7
7
9
.35
‐
VitaminD
supplementation,%
3
1
3
4
.63
‐
61(17)
93(12)
100(14)
102(15)
1.68(0.50)
2.47(0.41)
2.77(0.50)
56(7)
73(5)
75(5)
77(5)
<.001 AB,AC,AD,BC,BD,
CD
48(26‐64)
48(27‐60)
25(10‐41)
0(0)
<.001 AC,AD,BC,BD,CD
Lungfunctionandsmoking
FEV1,%predc
FEV1,L
FEV1/FVC,ratio
Packyearsb
<.001 AB,AC,AD,BC,BD,
CD
2.78(0.49) <.001 AB,AC,AD,BC,BD
Systemicinflammatorymarkers
IL‐6,pg/mlb
2.2(1.6‐3.6) 2.5(1.6‐3.9) 1.7(1.2‐2.7) 1.7(1.2‐2.3) <.001
AC,AD,BD
CRP,µg/ml 2.0(1.2‐3.5) 1.5(0.9‐3.3) 1.3(0.9‐2.4) 1.3(0.9‐2.2) <.001
AC,AD
TNF‐α,pg/mlb
3.1(2.4‐4.3) 3.2(2.3‐4.4) 3.3(2.6‐4.3) 3.1(2.4‐4.0)
b
.23
‐
Abbreviations:CRP,C‐reactiveprotein;IL,interleukin;FEV1,forcedexpiratoryvolumein1s;FVC,forced
vital capacity; SPPB, short physical performance battery; TNF‐α, tumor necrosis factor‐α. Data are
mean (SD) unless specified otherwise. aCorrected for multiple testing. bMedian (IQR). cFrom reference
equations[21].
the adjusted intercepts from the multilevel models (Figures 3 and 4 and the
corresponding Supplemental Tables 2 and 3) showed similar findings to the
baseline comparisons in Figure 2. While at baseline large differences were
observedinbodycompositionandphysicalfunctioning,theratesoflongitudinal
change of weight, fat mass, lean mass and strength were comparable between
OLD persons and never‐smoking controls (all P>.05), except for the change in
SPPBinOLDmen(P=.01)andBMCinOLDwomen(P=.02).Insmokingcontrols,
onlyleanmassdeclinedfasterinwomen(P=.03)andBMCfasterinmen(P=.02)
24
CHAPTER2
Table2.BaselinecharacteristicsofwomenintheHealthABCStudyaccordingto
obstructivelungdisease(OLD)statusandsmokingstatus.
OLD(n=110)
Smoking
controls
(n=83)
Formerly
smoking
controls
(n=332)
Never–
smoking
controls
(n=590)
A
B
C
D
72.8(2.8)
73.7(3.0)
73.3(2.8)
73.5(2.9)
.06
‐
62
37
56
59
.001
AB,BC,BD
51
.11
P
Significant
post‐hoc
comparisonsa
Demographics
Age,y
Race,%white
Site,%Memphis
42
43
46
Physicalactivity,
kcal/kg/wkb
67(36‐91)
66(36‐107)
65(40‐97)
74(45‐113) .006
‐
CD
Co‐morbidity,steroiduse,calciumandvitaminDsupplementation
Depression,%
12
16
11
11
.70
‐
Cardiovascular
disease,%
18
27
22
17
.42
‐
Diabetes,%
10
15
11
13
.63
‐
Oralsteroiduse,%
9
2
3
2
.001
AD
Calcium
supplementation,%
27
24
29
32
.35
‐
VitaminD
supplementation,%
9
7
13
15
.10
‐
Lungfunctionandsmoking
FEV1,%predc
FEV1,L
FEV1/FVC,ratio
Packyearsb
65(19)
97(19)
101(18)
1.24(0.38)
1.66(0.33)
1.87(0.38)
104(18)
<.001 AB,AC,AD,BD,CD
1.95(0.36) <.001 AB,AC,AD,BC,BD,
CD
58(7)
75(6)
76(5)
77(5)
<.001 AB,AC,AD,BD,CD
20(0‐51)
29(19‐50)
15(5‐38)
0(0)
<.001 AB,AD,BC,BD,CD
Systemicinflammatorymarkers
IL‐6,pg/mlb
1.9(1.3‐2.9) 1.9(1.3‐2.8) 1.7(1.2‐2.5)
1.6(1.1‐2.5) .021
BC
CRP,µg/mlb
2.1(1.1‐3.7) 2.0(1.2‐4.0) 2.2(1.3‐3.8)
1.6(1.0‐3.1) <.001
CD
TNF‐α,pg/mlb
2.9(2.2‐3.8) 3.1(2.2‐4.0) 3.0(2.4‐3.9)
3.1(2.3‐4.0)
.63
‐
Abbreviations:CRP,C‐reactiveprotein;IL,interleukin;FEV1,forcedexpiratoryvolumein1s;FVC,forced
vital capacity; SPPB, short physical performance battery; TNF‐α, tumor necrosis factor‐α. Data are
mean (SD) unless specified otherwise. aCorrected for multiple testing. bMedian (IQR). cFrom reference
equations[21].
comparedtonever‐smokingcontrols,whilethedeclinesinweightandfatmass
werecomparable(P>.05).
Further adjustment for baseline systemic cytokines resulted in a modest
attenuation of several outcome measures (Supplemental Tables 2 and 3). In
particular,thelowerbaselineandfasterdeclineofBMCinsmokingcontrolmen
(andtoalesserextentinwomen)wasexplainedbythecytokinesfor26%and
25%,respectively.ThelowerquadricepsstrengthinmenandwomenwithOLD
wasexplainedfor12%and18%bythecytokines,respectively.
25
Figure2.Baselinecharacteristicsofbodycompositionandphysicalfunctioning
inmenandwomen.
26
CHAPTER2
Figure2.continued.
Abbreviations:OLD,obstructivelungdisease;SPPB,shortphysicalperformancebattery.Barsrepresent
themeansanderrorbarsrepresenttheSE.*P<.05.
In additional analyses, pack years were added to model 1, which did not
result in any major differences. Compared to the never‐smoking controls, the
formerly‐smoking controls, smoking controls, and OLD persons had slightly
higher probabilities of being missing along the seven years of follow‐up.
However, thetrajectoriesof the outcomes essentially remainedthe same when
we adjusted for missingness (not tabulated). The percentages of participants
being alive at the year‐8 visit were 62%, 66%, 77% and 85% for the OLD
27
Figure 3. Longitudinal course of body composition and physical functioning
accordingtoobstructivelungdiseaseandsmokingstatusinmen.a
persons, smoking controls, formerly smoking controls and never‐smoking
controls,respectively.
Discussion
In the current study, we have shown striking similarities in body composition
and physical functioning between older adult men and women with mild‐to‐
moderate OLD and smokers with normal lung function. These were
characterized by a markedly lower body weight, lean mass, fat mass, BMC and
quadricepsstrengthcomparedtonever‐smokingcontrols.Interestingly,because
longitudinal trajectories in the 8th decade of life were to a large extent
comparable to those of their never‐smoking controls, the large baseline
differences persisted over a period of 7 years follow up. This suggests that a
common insult related to smoking induced a divergent pattern earlier in life,
whichmaynotuniformlypersistintoold‐age.
28
CHAPTER2
Figure3.continued.a
Abbreviations: OLD, obstructive lung disease; SPPB, short physical performance battery. aLines
representthemeanpredictedvaluesadjustedfortime,age,site,race,diabetes,cardiovasculardisease,
depression, physical activity, oral steroid use, calcium supplementation, vitamin D supplementation,
time2, age⨯time, race⨯time, site⨯time, diabetes⨯time, cardiovascular disease⨯time, depression⨯time
and physical activity⨯time, oral steroid use⨯time, calcium supplementation⨯time and vitamin D
supplementation⨯time. #Intercept significantly different from never‐smoking controls, P<.05. ¶Slope
significantlydifferentfromnever‐smokingcontrols,P<.05.
TheuniquedesignoftheHealthABCStudyenabledustoinvestigateforthe
firsttimethelongitudinalpatternandprogressionofbodycompositionchanges
andthedeclineinphysicalfunctioningsimultaneouslyinpersonswithOLDand
smokerswithoutOLDincomparisontoagingeffectsinanever‐smokingcontrol
groupwithoutOLD.Furthermore,bydesignoftheHealthABCStudy,noneofthe
participants had functional limitations at baseline, and likely as a result of this
selection, physical activity and prevalence of co‐morbidity were comparable
between the study groups. The Health ABC Study was therefore particularly of
interest to examine the effects of smoking and OLD on body composition and
physicalfunctioningwithaging.Ourresultsmay,however,notbegeneralizedto
allolderadultsnortoclinicalCOPDpatientssincestudyparticipantswere70‐79
years old and well‐functioning at baseline, and the OLD group was defined by
lungfunctionratherthanaclinicalCOPDdiagnosis.Therefore,itcannotberuled
29
Figure 4. Longitudinal course of body composition and physical functioning
accordingtoobstructivelungdiseaseandsmokingstatusinwomen.a
out that our data might underestimate “real” changes. Another strength of our
studyisthatwecarriedoutasensitivityanalysistoestimatethepotentialeffect
of missingobservationson the observed trajectories.This effect was estimated
to be small which, however, does not rule out that survival bias may have
obscured our results. While we were unable to examine the effect of smoking
statuswiththeOLDpersonsduetosamplesizeconsiderations,resultsbetween
groupsremainedsimilarafterfurtheradjustmentforpackyears,suggestingthat
theobserveddifferencesareindependentofthequantityofsmoking.
WefoundthatthedeclineinFEV1waslowerinOLDmencomparedtotheir
never‐smoking controls. Because this data was generated from a relatively low
numberoffollow‐uplungfunctionmeasureswearecarefulinitsinterpretation,
but it suggests that longitudinal body composition and physical functioning
trajectories in OLD persons in our study are not related to a faster decline in
FEV1.
In OLD men, but not in OLD women, we found an accelerated decline in
SPPB.Thisgenderdifferencemaybeexplainedbyagreaterrelativedifferencein
leg lean mass and quadriceps strength between OLD persons and never‐
smoking controls in men than in women. This is in line with a previous report
30
CHAPTER2
Figure4.continued.a
Abbreviations: OLD, obstructive lung disease; SPPB, short physical performance battery. aLines
representthemeanpredictedvaluesadjustedfortime,age,site,race,diabetes,cardiovasculardisease,
depression, physical activity, oral steroid use, calcium supplementation, vitamin D supplementation,
time2, age⨯time, race⨯time, site⨯time, diabetes⨯time, cardiovascular disease⨯time, depression⨯time
and physical activity⨯time, oral steroid use⨯time, calcium supplementation⨯time and vitamin D
supplementation⨯time. #Intercept significantly different from never‐smoking controls, P<.05. ¶Slope
significantlydifferentfromnever‐smokingcontrols,P<.05.
showingthattherelationbetweenalowerfat‐freemassandphysicaldisability
wasmorepronouncedinmenwithCOPD[29].AlowerSPPBhasbeenstrongly
related to subsequent disability and mortality [24, 30], and may thus be of
clinical importance in particularly in men with OLD. Notably, despite a similar
pattern of daily physical activity, quadriceps strength was much better able to
discriminate between OLD persons, smoking controls and never‐smoking
controls than hand grip strength. Smoking control women had an accelerated
decline in lean mass compared to never‐smoking controls, which confirms
previous cross‐sectional data of a sarcopenic phenotype in older smoking
women[31].
Systemic inflammation has been considered a key player in the
extrapulmonary manifestations of COPD [5], and has been linked to muscle
atrophy [32] and bone demineralization [33]. In the present study we indeed
31
confirmed the presence of mildly elevated circulating cytokines in particularly
OLD men but also in smoking controls. After further adjustment for systemic
inflammatory markers we found little attenuation for the majority of outcome
measures, but systemic inflammatory markers did explain 11‐26% of the
observedbaselinedifferencesinBMC,andexplained25%oftheacceleratedBMC
declineinsmokingcontrolmen.
The observation that smoking cessation is associated with weight gain is
well‐known [34]. While this effect is mainly considered a relatively short‐term
effect(months)[35],formersmokersinourstudy–whoabstainedfromsmoking
for~25years–stillshowedatrendtowardshigherbaselineweight,whichwas
mainly due to higher fat mass. Apart from fat mass, body composition and
physical functioning in former smokers were comparable to that of never‐
smokers,suggestingahighrecoverypotentialfromsmoking‐inducedlosses.As
higherbodyweightandleanmasshavebeenassociatedwithbettersurvivalin
COPD [36], it needs to be determined whether this recovery potential is as
functional in the state of COPD. We found higher CRP in formerly smoking
women, and borderline significantly higher TNF‐α in formerly smoking men,
which may be related to metabolic effects associated with fat abundance or
alteredfatdistributioninobesity.
Although it was not part of our analyses, our data raise the possibility that
low body and muscle mass may increase the risk of COPD development. In
support of this, recent data from a large group of never‐smokers from the
Burden of Obstructive Lung Disease Study showed that the risk of COPD was
significantlyincreasedinthosewithaBMI<20kg/m2,especiallyinwomen[37].
In the case of smoking‐induced COPD, however, it remains to be investigated
whetheracompromisedbodycompositionrelatedtosmokingcontributestothe
riskofCOPDdevelopment.Ourobservationthatwhileoursmokingcontrolshad
comparable pack years smoked as OLD persons and had a similarly
compromised body composition but no airflow obstruction, indicates that it is
notsolelylowbodyweightthatwouldincreasestheriskofCOPDdevelopment,
but suggests that other factors such as genetic susceptibility to for example
emphysemaandassociatedwastingareimportant.
In conclusion, our study shows that smoking is not only an important risk
factorinOLDetiology,butalsoadverselyaffectsbodycompositionandphysical
functioning. The individual contributions of impaired weight gain and
accelerated weight loss and a potential role for cytokinaemic stress associated
withsmokingremaintobestudiedincomparablefuturelongitudinalstudiesin
middle‐agedpersons.
32
CHAPTER2
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
GooneratneNS,PatelNP,CorcoranA.ChronicObstructivePulmonaryDiseasediagnosisandmanagementin
olderadults.JAmGeriatrSoc.2010;58:1153‐62.
Celli BR, MacNee W. Standards for the diagnosis and treatment of patients with COPD: a summary of the
ATS/ERSpositionpaper.EurRespirJ.2004;23(6):932‐46.
ChatilaWM,ThomashowBM,MinaiOA,CrinerGJ,MakeBJ.Comorbiditiesinchronicobstructivepulmonary
disease.ProcAmThoracSoc.2008;5(4):549‐55.
Gan WQ, Man SF, Senthilselvan A, Sin DD. Association between chronic obstructive pulmonary disease and
systemicinflammation:asystematicreviewandameta‐analysis.Thorax.2004;59(7):574‐80.
WoutersEF,ReynaertNL,DentenerMA,VernooyJH.Systemicandlocalinflammationinasthmaandchronic
obstructivepulmonarydisease:isthereaconnection?ProcAmThoracSoc.2009;6(8):638‐47.
VestboJ,PrescottE,AlmdalT,DahlM,NordestgaardBG,AndersenT,SorensenTI,LangeP.Bodymass,fat‐free
bodymass,andprognosisinpatientswithchronicobstructivepulmonarydiseasefromarandompopulation
sample:findingsfromtheCopenhagenCityHeartStudy.AmJRespirCritCareMed.2006;173(1):79‐83.
EisnerMD,IribarrenC,YelinEH,SidneyS,KatzPP,AckersonL,LathonP,TolstykhI,OmachiT,BylN,etal.
Pulmonary function and the risk of functional limitation in chronic obstructive pulmonary disease. Am J
Epidemiol.2008;167(9):1090‐101.
2007;62(11):944‐9.
Graat‐VerboomL,WoutersEF,SmeenkFW,vandenBorneBE,LundeR,SpruitMA.Currentstatusofresearch
onosteoporosisinCOPD:asystematicreview.EurRespirJ.2009;34(1):209‐18.
YendeS,WatererGW,TolleyEA,NewmanAB,BauerDC,TaaffeDR,JensenR,CrapoR,RubinS,NevittM,etal.
Inflammatorymarkersareassociatedwithventilatorylimitationandmuscledysfunctioninobstructivelung
diseaseinwellfunctioningelderlysubjects.Thorax.2006;61(1):10‐6.
Scanlon PD, Connett JE, Wise RA, Tashkin DP, Madhok T, Skeans M, Carpenter PC, Bailey WC, Buist AS,
EichenhornM,etal.LossofbonedensitywithinhaledtriamcinoloneinLungHealthStudyII.AmJRespirCrit
CareMed.2004;170(12):1302‐9.
OgaT,NishimuraK,TsukinoM,SatoS,HajiroT,MishimaM.Exercisecapacitydeteriorationinpatientswith
COPD:longitudinalevaluationover5years.Chest.2005;128(1):62‐9.
OgaT,NishimuraK,TsukinoM,SatoS,HajiroT,MishimaM.Longitudinaldeteriorationsinpatientreported
outcomesinpatientswithCOPD.RespirMed.2007;101(1):146‐53.
HabrakenJM,vanderWalWM,TerRietG,WeersinkEJ,TobenF,BindelsPJ.Health‐relatedqualityoflifeand
functionalstatusinend‐stageCOPD:alongitudinalstudy.EurRespirJ.2011;37(2):280‐8.
KosterA,VisserM,SimonsickEM,YuB,AllisonDB,NewmanAB,vanEijkJT,SchwartzAV,SatterfieldS,Harris
TB. Association between fitness and changes in body composition and muscle strength. J Am Geriatr Soc.
2010;58(2):219‐26.
Fulop T, Larbi A, Witkowski JM, McElhaney J, Loeb M, Mitnitski A, Pawelec G. Aging, frailty and age‐related
diseases.Biogerontology.2010;11(5):547‐63.
Montes de Oca M,Loeb E, Torres SH, De SanctisJ, Hernandez N, Talamo C. Peripheral muscle alterationsin
non‐COPDsmokers.Chest.2008;133(1):13‐8.
Wust RC, Morse CI, de Haan A, Rittweger J, Jones DA, Degens H. Skeletal muscle properties and fatigue
resistanceinrelationtosmokinghistory.EurJApplPhysiol.2008;104(1):103‐10.
Tanni SE, Pelegrino NR, Angeleli AY, Correa C, Godoy I. Smoking status and tumor necrosis factor‐alpha
mediatedsystemicinflammationinCOPDpatients.JInflamm(Lond).2010;7:29.
LevitzkyYS,GuoCY,RongJ,LarsonMG,WalterRE,KeaneyJF,Jr.,SutherlandPA,VasanA,LipinskaI,EvansJC,
et al. Relation of smoking status to a panel of inflammatory markers: the framingham offspring.
Atherosclerosis.2008;201(1):217‐24.
Hankinson JL, Odencrantz JR, Fedan KB. Spirometric reference values from a sample of the general U.S.
population.AmJRespirCritCareMed.1999;159(1):179‐87.
MehrotraN,FreireAX,BauerDC,HarrisTB,NewmanAB,KritchevskySB,MeibohmB.Predictorsofmortality
inelderlysubjectswithobstructiveairwaydisease:thePILEscore.AnnEpidemiol.2010;20(3):223‐32.
Guralnik JM, Simonsick EM, Ferrucci L, Glynn RJ, Berkman LF, Blazer DG, Scher PA, Wallace RB. A short
physicalperformancebatteryassessinglowerextremityfunction:associationwithself‐reporteddisabilityand
predictionofmortalityandnursinghomeadmission.JGerontol.1994;49(2):M85‐94.
GuralnikJM,FerrucciL,SimonsickEM,SaliveME,WallaceRB.Lower‐extremityfunctioninpersonsoverthe
ageof70yearsasapredictorofsubsequentdisability.NEnglJMed.1995;332(9):556‐61.
KosterA,StenholmS,AlleyDE,KimLJ,SimonsickEM,KanayaAM,VisserM,HoustonDK,NicklasBJ,Tylavsky
FA, et al. Body Fat Distribution and Inflammation Among Obese Older Adults With and Without Metabolic
Syndrome.Obesity(SilverSpring).2010;18(12):2354‐61.
KosterA,PatelKV,VisserM,vanEijkJT,KanayaAM,deRekeneireN,NewmanAB,TylavskyFA,Kritchevsky
SB,HarrisTB.Jointeffectsofadiposityandphysicalactivityonincidentmobilitylimitationinolderadults.J
AmGeriatrSoc.2008;56(4):636‐43.
AinsworthBE,HaskellWL,WhittMC,IrwinML,SwartzAM,StrathSJ,O'BrienWL,BassettDR,Jr.,SchmitzKH,
Emplaincourt PO, et al. Compendium of physical activities: an update of activity codes and MET intensities.
MedSciSportsExerc.2000;32(9Suppl):S498‐504.
33
28. Ghosh P, Tu W. Assessing Sexual Attitudes and Behaviors of Young Women: A Joint Model with Nonlinear
TimeEffects,TimeVaryingCovariates,andDropouts.JAmStatAssoc.2008;103(484):1496‐507.
29. Schols AM, Broekhuizen R, Weling‐Scheepers CA, Wouters EF. Body composition and mortality in chronic
obstructivepulmonarydisease.AmJClinNutr.2005;82(1):53‐9.
30. MelzerD,LanTY,GuralnikJM.Thepredictivevalidityformortalityoftheindexofmobility‐relatedlimitation‐‐
resultsfromtheEPESEstudy.AgeAgeing.2003;32(6):619‐25.
31. CastilloEM,Goodman‐GruenD,Kritz‐SilversteinD,MortonDJ,WingardDL,Barrett‐ConnorE.Sarcopeniain
elderlymenandwomen:theRanchoBernardostudy.AmJPrevMed.2003;25(3):226‐31.
32. LangenRC,ScholsAM,KeldersMC,vanderVeldenJL,WoutersEF,Janssen‐HeiningerYM.Musclewastingand
impaired muscle regeneration in a murine model of chronic pulmonary inflammation. Am J Respir Cell Mol
Biol.2006;35(6):689‐96.
33. McLeanRR.Proinflammatorycytokinesandosteoporosis.CurrOsteoporosRep.2009;7(4):134‐9.
34. Filozof C, Fernandez Pinilla MC, Fernandez‐Cruz A. Smoking cessation and weight gain. Obes Rev.
2004;5(2):95‐103.
35. CropseyKL,McClureLA,JacksonDO,VillalobosGC,WeaverMF,StitzerML.TheImpactofQuittingSmokingon
Weight Among Women Prisoners Participating in a Smoking Cessation Intervention. Am J Public Health.
2010;100(8):1442‐8.
36. Schols AM, Slangen J, Volovics L, Wouters EF. Weight loss is a reversible factor in the prognosis of chronic
37. LamprechtB,McBurnieMA,VollmerWM,GudmundssonG,WelteT,Nizankowska‐MogilnickaE,StudnickaM,
BatemanE,AntoJM,BurneyP,etal.COPDinNeverSmokers:ResultsFromthePopulation‐BasedBurdenof
ObstructiveLungDiseaseStudy.Chest.2011;139(4):752‐63.
34
CHAPTER2
Supplementalmaterial
SupplementalTable1.Declineinlungfunctioninn=1658participantswithat
leasttwovalidlungfunctionmeasurementsintime(mean[SD]follow‐upperiod
of6.2[1.3]yearsincluding2.6[0.5]validmeasurementsperparticipant).
OLD
Smoking
controls
Formerly
smoking
controls
Never‐
smoking
controls
P
Significant
post‐hoc
comparisonsa
A
B
C
D
91
48
406
240
‐38(6)
‐54(7)
‐55(3)
‐57(3)
.039
AD
Women
75
50
262
486
‐33(5)
‐39(5)
‐40(2)
‐40(2)
.45
‐
Men
n
FEV1(mL/yr)
n
FEV1(mL/yr)
Abbreviations:FEV1,forcedexpiratoryvolumein1s;OLD,obstructivelungdisease.Dataaremeans(SE).
aCorrectedformultipletesting.
35
SupplementalTable2.Interceptsandslopesofbodycompositionandphysical
functioninginmen.
Model1
Weight(kg)
Intercept
P
Model2
Slope
P
Intercept
P
Slope
P
Never‐smokingcontrols
Ref
Ref
Ref
Ref
Formerlysmokingcontrols
1.66
.08
0.01
.88
1.19
.23
0.001
.92
Smokingcontrols
‐6.53
<.001
‐0.14
.37
‐6.34
.001
‐0.18
.26
OLD
‐5.08
<.001
‐0.07
.54
‐5.31
<.001
‐0.01
Fatmass(kg)
.94
Ref
Ref
Ref
Ref
1.34
.011
‐0.05
.36
1.12
.040
‐0.06
.29
Smokingcontrols
‐3.11
.001
‐0.08
.41
‐3.05
.003
‐0.11
.30
OLD
‐2.02
.006
‐0.10
.19
‐2.21
.005
‐0.06
.43
Ref
Ref
Ref
Ref
Leanmass(kg)
0.45
.37
‐0.01
.88
0.22
.68
‐0.00
.91
Smokingcontrols
‐3.15
.001
‐0.01
.86
‐3.01
.002
‐0.03
.69
OLD
‐2.83
<.001
‐0.07
.20
‐2.85
<.001
‐0.04
.46
Ref
Ref
Ref
Ref
Legleanmass(kg)
0.02
.92
‐0.01
.71
‐0.04
.82
‐0.00
.88
Smokingcontrols
‐1.27
<.001
‐0.02
.56
‐1.17
.001
‐0.03
.36
OLD
‐1.30
<.001
0.00
.88
‐1.26
<.001
0.01
.59
Ref
Ref
Ref
0.001
.20
Bonemineralcontent(kg)
Ref
‐0.007
.83
.27
‐0.012
.70
0.002
Smokingcontrols
‐0.143
.010
‐0.008 .017
‐0.106
.07
‐0.006
.08
OLD
‐0.176
<.001
0.002
‐0.170
<.001
0.003
.32
Ref
Ref
Ref
Ref
Quadricepsstrength(Nm)
.82
0.08
.98
‐0.12
.74
0.33
.90
‐0.04
.92
Smokingcontrols
‐7.18
.13
‐0.42
.56
‐4.43
.38
‐0.77
.30
OLD
‐14.21
<.001
0.90
.12
‐12.55
.001
0.76
.21
Abbreviations:OLD,obstructivelungdisease;Ref,reference.Theinterceptsdenotetheadjustedbaseline
differenceswiththenever‐smokingcontrols.Theslopesdenotetheadjusteddifferencesinyearlychange
with the never‐smoking controls. Model 1: adjusted for time, age, site, race, diabetes, cardiovascular
disease, depression, physical activity, oral steroid use, calcium supplementation, vitamin D
supplementation, time2,age⨯time, race⨯time,site⨯time,diabetes⨯time,cardiovascular disease⨯time,
depression⨯time, physical activity⨯time, oral steroid use⨯time, calcium supplementation⨯time,
vitamin D supplementation⨯time. Model 2: further adjustment for IL‐6, TNF‐α, CRP, IL‐6⨯time, TNF‐
α⨯timeandCRP⨯time.
36
CHAPTER2
SupplementalTable2.continued.
Model1
Intercept
Model2
P
Slope
P
Intercept
P
Slope
Ref
Ref
Ref
Ref
1.68
.13
‐0.07
.61
1.51
.19
‐0.03
.82
Smokingcontrols
‐3.65
.07
0.09
.71
‐3.42
.11
0.16
.54
OLD
‐0.57
.71
‐0.13
.71
0.01
.99
‐0.14
.51
Ref
Ref
Ref
Ref
Handgripstrength(kg)
SPPB(0‐12)
P
‐0.09
.43
‐0.00
.94
‐0.11
.33
0.01
.71
Smokingcontrols
‐0.26
.21
‐0.04
.53
‐0.27
.22
‐0.01
.87
OLD
‐0.22
.16
‐0.12
.01
‐0.14
.40
‐0.14
.01
Abbreviations:OLD,obstructivelungdisease;Ref,reference;SPPB,shortphysicalperformancebattery.
The intercepts denote the adjusted baseline differences with the never‐smoking controls. The slopes
denotetheadjusteddifferencesinyearlychangewiththenever‐smokingcontrols.Model1:adjustedfor
time, age, site, race, diabetes, cardiovascular disease, depression, physical activity, oral steroid use,
calcium supplementation, vitamin D supplementation, time2, age⨯time, race⨯time, site⨯time,
diabetes⨯time, cardiovascular disease⨯time, depression⨯time, physical activity⨯time, oral steroid
use⨯time, calcium supplementation⨯time, vitamin D supplementation⨯time. Model 2: further
adjustmentforIL‐6,TNF‐α,CRP,IL‐6⨯time,TNF‐α⨯timeandCRP⨯time.
37
SupplementalTable3.Interceptsandslopesofbodycompositionandphysical
functioninginwomen.
Model1
Model2
Intercept
P
Slope
P
Intercept
P
Slope
Ref
Ref
Ref
Ref
Weight(kg)
P
1.29
.17
‐0.08
.25
1.55
.11
‐0.04
.58
Smokingcontrols
‐4.93
.003
‐0.14
.29
‐5.56
.001
‐0.07
.61
OLD
‐4.29
.003
‐0.19
.09
‐3.51
.019
‐0.17
.15
Ref
Ref
Ref
Ref
Fatmass(kg)
0.86
.17
‐0.00
.99
0.89
.16
0.03
.51
Smokingcontrols
‐3.55
.001
0.03
.72
‐4.13
<.001
0.09
.32
OLD
‐2.62
.006
‐0.04
.57
‐2.23
.023
‐0.02
.81
Ref
Ref
Ref
Ref
‐0.01
.55
‐0.01
.59
Leanmass(kg)
0.33
.39
0.55
.17
Smokingcontrols
‐1.38
.039
‐0.10 .028 ‐1.43
.039
‐0.10 .049
OLD
‐1.72
.003
‐0.03
.40
‐1.34
.030
‐0.02
Ref
Ref
Ref
Ref
‐0.02
.07
‐0.02
.09
Legleanmass(kg)
.56
0.12
.42
0.21
.19
Smokingcontrols
‐0.62
.018
‐0.07 .001 ‐0.66
.016
‐0.06 .002
OLD
‐0.71
.002
‐0.01
.72
‐0.58
.017
‐0.00
Ref
Ref
Ref
.93
.79
0.001
.48
Bonemineralcontent(kg)
Ref
.94
0.002
0.002
.19
0.006
Smokingcontrols
‐0.121
.003 ‐0.001
.68
‐0.108
.010 ‐0.002 .32
OLD
‐0.093
.008 ‐0.004 .024 ‐0.083
.027 ‐0.005 .017
Quadricepsstrength(Nm)
Ref
Ref
Ref
Ref
.77
‐0.66
.69
0.13
.70
‐0.42
.82
0.11
Smokingcontrols
‐6.05
.048
0.09
.87
‐5.27
.10
‐0.09
.89
OLD
‐7.57
.003
0.12
.82
‐6.21
.025
0.06
.92
Abbreviations:OLD,obstructivelungdisease;Ref,reference.Theinterceptsdenotetheadjustedbaseline
differenceswiththenever‐smokingcontrols.Theslopesdenotetheadjusteddifferencesinyearlychange
with the never‐smoking controls. Model 1: adjusted for time, age, site, race, diabetes, cardiovascular
disease, depression, physical activity, oral steroid use, calcium supplementation, vitamin D
supplementation, time2,age⨯time, race⨯time,site⨯time,diabetes⨯time,cardiovascular disease⨯time,
depression⨯time, physical activity⨯time, oral steroid use⨯time, calcium supplementation⨯time,
vitamin D supplementation⨯time. Model 2: further adjustment for IL‐6, TNF‐α, CRP, IL‐6⨯time, TNF‐
α⨯timeandCRP⨯time.
38
CHAPTER2
SupplementalTable3.continued.
Model1
Intercept
Model2
P
Slope
P
Intercept
P
Slope
Ref
Ref
Ref
Ref
0.77
.29
‐0.06
.55
0.42
.58
‐0.04
.71
Smokingcontrols
‐1.53
.24
0.02
.93
‐1.35
.31
0.06
.78
OLD
‐1.64
.14
0.12
.44
‐1.57
.19
0.14
.42
Ref
Ref
Ref
Ref
Handgripstrength(kg)
SPPB(0‐12)
P
0.00
.99
‐0.02
.61
‐0.01
.96
‐0.04
.29
Smokingcontrols
‐0.20
.33
0.03
.61
‐0.24
.26
0.03
.65
OLD
‐0.33
.06
‐0.04
.40
‐0.34
.07
‐0.09
.10
Abbreviations:OLD,obstructivelungdisease;Ref,reference;SPPB,shortphysicalperformancebattery.
The intercepts denote the adjusted baseline differences with the never‐smoking controls. The slopes
denotetheadjusteddifferencesinyearlychangewiththenever‐smokingcontrols.Model1:adjustedfor
time, age, site, race, diabetes, cardiovascular disease, depression, physical activity, oral steroid use,
calcium supplementation, vitamin D supplementation, time2, age⨯time, race⨯time, site⨯time,
diabetes⨯time, cardiovascular disease⨯time, depression⨯time, physical activity⨯time, oral steroid
use⨯time, calcium supplementation⨯time, vitamin D supplementation⨯time. Model 2: further
adjustmentforIL‐6,TNF‐α,CRP,IL‐6⨯time,TNF‐α⨯timeandCRP⨯time.
39
CHAPTER3
Loss of quadriceps muscle oxidative
phenotypeanddecreasedendurancein
patientswithmild‐to‐moderateCOPD
BramvandenBorst,IlseG.M.Slot,ValéryA.C.V.Hellwig,BettineA.H.
Vosse, Marco C.J.M. Kelders, Esther Barreiro, Annemie M.W.J. Schols,
HarryR.Gosker
JournalofAppliedPhysiology2012Jul19(Epubaheadofprint)
41
SKELETALMUSCLEFATIGUEINEARLYCOPD
Abstract
Background:Beingwell‐establishedinadvancedchronicobstructivepulmonary
disease (COPD), skeletal muscle dysfunction and its underlying pathology have
been scarcely investigated in patients with mild‐to‐moderate airflow
obstruction. We hypothesized that a loss of oxidative phenotype (oxphen),
associated with decreased endurance, is already present in skeletal muscle of
mild‐to‐moderateCOPDpatients.
Methods: In quadriceps muscle biopsies from n=29 COPD patients (forced
expiratory volume in 1s [FEV1] 58±16 %pred, body mass index [BMI] 26±4
kg/m2)andn=15controls(BMI25±3kg/m2)weassessedfibertypedistribution,
fiber cross‐sectional areas (CSA), oxidative and glycolytic gene expression,
OXPHOSproteinlevels,metabolicenzymeactivity,andlevelsofoxidativestress
markers. Quadriceps function was assessed by isokinetic dynamometry, body
composition by dual‐energy X‐ray absorptiometry, exercise capacity by an
incrementalloadtest,andphysicalactivitylevelbyaccelerometry.
Results: Compared to controls, patients had comparable fat‐free mass index
(FFMI), quadriceps strength and fiber CSA, but quadriceps endurance was
decreasedby29%(P=.002).Patientshadaclearlossofmuscleoxphen:afiber
typeI‐to‐IIshift,decreasedlevelsofOXPHOScomplexesIVandVsubunits(47%
and 31% respectively, P<.05), a decreased ratio of 3‐hydroxyacyl‐CoA
dehydrogenase/phosphofructokinase(PFK)enzymeactivities(38%,P<.05),and
decreased peroxisome proliferator‐activated receptor‐γ coactivator‐1α (40%;
P<.001)versusincreasedPFK(67%;P<.001)geneexpressionlevels.Withinthe
patients, markers of oxphen were significantly positively correlated with
quadricepsenduranceandinverselywiththeincreaseinplasmalactaterelative
to work rate during the incremental test. Levels of protein carbonylation,
tyrosine nitration and malondialdehyde protein adducts were comparable
betweenpatientsandcontrols.However,amongpatientsoxidativestresslevels
significantly correlated inversely with markers of oxphen and quadriceps
endurance.
Conclusion: Reduced muscle endurance associated with underlying loss of
muscle oxphen is already present in patients with mild‐to‐moderate COPD
withoutmusclewasting.
42
CHAPTER3
Introduction
Skeletal muscle dysfunction is a hallmark of advanced chronic obstructive
pulmonary disease (COPD) and significantly contributes to decreased exercise
capacityandpoorqualityoflife[1,2].Peripheralskeletalmusclewastinganda
loss of oxidative phenotype (oxphen) are the two major myopathological
findingsrecognizedinadvancedCOPDpatients,beingassociatedwithdecreased
muscle strength [3] and endurance [4], respectively. Important to note is that
studiesinvestigatingskeletalmuscledysfunctionanditsunderlyingpathologyin
COPDhavebeenperformedalmostexclusivelyinpatientswithGOLDstages3‐4
withsignificantmusclewasting.IntheInterdisciplinaryCommunity‐basedCOPD
managementtrial,VanWeteringetal.[5]haverecentlyshownthatpatientswith
onaveragemild‐to‐moderateCOPD(meanforcedexpiratoryvolumein1s[FEV1]
60%ofpredicted)respondedwelltoalifestyleinterventionprogramintermsof
improvementincycleendurancecapacityandhealthstatus.Musclestrengthand
muscle mass of that study group were within the normal range, but no
informationwasavailableregardingperipheralmuscleenduranceandmarkers
ofskeletalmuscleoxphen.
LossofperipheralmuscleoxpheninadvancedCOPDincludesafibertypeI‐
to‐II shift and reduced activities of enzymes involved in oxidative energy
metabolism [4, 6, 7]. In addition, Remels et al. [8] previously showed that key
oxphenregulators(peroxisomeproliferator‐activatedreceptors[PPARs],PPAR‐
γ coactivator‐1α [PGC‐1α], mitochondrial transcription factor A [TFAM]) were
reduced in muscles of patients with advanced COPD. Moreover, loss of muscle
oxphen may render the muscle more susceptible to oxidative stress [9‐11].
Oxidativestressisthoughttobeanimportantplayerinskeletalmusclewasting
and dysfunction in COPD [12, 13]. However, data on muscle oxphen and
oxidativestressinCOPDpatientswithmilderdegreesofairflowobstructionare
scarce.
In the current study we compared markers of oxphen measured in
quadriceps muscle biopsies from patients with on average mild‐to‐moderate
COPD and age‐, sex‐, and body mass index (BMI)‐matched healthy controls. In
addition, we explored associations between muscle metabolic profile and
quadricepsfunction,exercisecapacity,physicalactivitylevelandskeletalmuscle
oxidative stress. We hypothesized that a loss of oxphen, associated with
decreasedendurance,isalreadypresentinskeletalmuscleofmild‐to‐moderate
COPDpatients.
43
Methods
Subjectsandstudydesign
The study population comprised n=29 clinically stable mild‐to‐moderate COPD
patientsandn=15healthycontrols.Patientswererecruitedfromtheoutpatient
clinic of the Maastricht University Medical Center+ (MUMC+, Maastricht, The
Netherlands) and via advertisements in local newspapers. Patients were
excluded in case of long‐term oxygen therapy, oral prednisolone use, acute
exacerbationwithhospitaladmissioninthepasteightweeksandrehabilitation
inthepastsixmonths.Patientswithknownco‐morbiditypotentiallyinterfering
with study outcome parameters were carefully excluded. These included
diabetes,recentcardiovascularevents,inflammatoryboweldisease,obstructive
sleep apnea, thyroid disease, and cancer. Healthy controls were recruited via
advertising in local newspapers. The absence of these diseases in the healthy
subjects was verified through history‐taking by a physician, and pulmonary
function tests verified absence of airflow limitation in these subjects. Care was
taken to select a group of healthy controls with similar age, BMI and sex
distributionastheCOPDpatients,regardlessoftheirdailyphysicalactivitylevel.
Writteninformedconsentwasobtainedfromallsubjectsandtheethicalreview
boardoftheMUMC+approvedthestudy(08‐2‐059).Thetrialwasregisteredat
www.trialregister.nlasNTR1402.
On day 1, body composition and quadriceps function were assessed as
described below, and subjects started wearing an accelerometer for 6 days in
ordertoobjectivelyquantifyphysicalactivitylevel.Onday7,theaccelerometry
data was retrieved, quadriceps tissue samples were obtained, pulmonary
functionwasassessed,andanincrementalloadtestwasperformed.
Pulmonaryfunction,smokingstatusandexercisecapacity
Pulmonary function testing included forced spirometry and single breath
diffusion capacity measurement (Masterlab, Jaeger, Würzburg, Germany).
Instrumentswerecalibratedtwiceaday.Allvaluesobtainedwereexpressedasa
percentageofreferencevalues[14].Smokingstatuswasbasedonself‐reportand
those who had smoked or were current smokers were designated as ever‐
smokers. All subjects performed an incremental load cycling test to determine
peakoxygenuptake(VO2peak)andpeakload(Wpeak)aspreviouslydescribed
[15]. Arterial punctures of the radial artery at rest and at VO2 peak were
availablefromn=19COPDpatientsandn=11healthysubjects.Arterialbloodgas
analyses and lactate concentrations were determined (Blood gas analyzer 865,
ChironDiagnostics,Emeryville,CA).
44
CHAPTER3
Anthropometryandbodycomposition
Heightinmetersandweightinkilogramswereassessedonastandardscale.BMI
wascalculatedasweight/height2.Whole‐bodydual‐energyX‐rayabsorptiometry
was performed to assess body composition as described [16]. Fat‐free mass
index(FFMI)wascalculatedasfat‐freemass/height2.Theprevalenceofmuscle
depletionwasexploredbyapplyingthecriteriafromScholsetal.[17](FFMI<16
kg/m2formenandFFMI<15kg/m2forwomen).
Quadricepsenduranceandstrength
Isokinetic muscle endurance and strength of the dominant knee extensor
(quadriceps muscle) were measured using a dynamometer (Biodex System,
Biodex Corporation, Shirley, New York, USA). Subjects were seated upright on
thechairofthedynamometerwithsupportoftheback.Atthelevelofthechest,
pelvisandthigh,subjectsweresecuredwithstraps.Thehipjointwasatanangle
between90‐100°offlexionduringtesting.Thetestconsistedofthirtysequential
volitionalmaximalcontractionsatanangularvelocityof90°/s,duringwhichthe
subjectwasstronglyencouraged.Maximalisokineticstrengthwasdefinedasthe
highest peak torque (in N·m) in this series of thirty. To determine isokinetic
quadricepsendurance,theproportionaldeclineinpeaktorques(relativetothe
highest peak torque) per repetition (%N·m/rep) was used as determined by
linearregression,asdescribedpreviously[18].Weusedthisslopeasameasure
of quadriceps endurance (i.e., a slope closer to zero indicates a better
endurance). Valid quadriceps endurance measures were available for n=24
COPDpatientsandn=13healthycontrols.
Dailyphysicalactivity
Physical activity (PA) was measured using a dual‐axis GT1M accelerometer
(ActiGraph™,LLC,FortWaltonBeach,FL,USA)forsixconsecutivedays(4week
daysand2weekenddays).Subjectswereinstructedtoweartheaccelerometer
during the time they were not asleep, except when showering or bathing [19].
Theaccelerometerwasfirmlyattachedtoanelasticbeltwornatthewaist.The
accelerometer registers PA in ‘counts’, which are the summation of the
accelerationsmeasuredduringaspecifiedtimeinterval(epoch),whichwasset
atoneminute.Countsrepresenttheintensityofactivityinthatepoch.Non‐wear
timewasdefinedas60consecutivecountsof0,inwhichuptotwoepochs<100
countswereallowed[20].Onlydayswith≥10hoursofweartimewereaccepted
as valid days [20]. The total amount of PA was expressed as the total counts
divided by the total wear time (counts/min). Cut‐off points for sedentary,
lifestyle and combined moderate‐to‐vigorous PA (MVPA) intensity levels were
defined as <100, 100‐759, ≥760 counts/min, respectively [20, 21]. The time
45
spentineachcategoryofintensitywaspresentedasapercentageoftotalwear
time. Furthermore, we analyzed the number, duration and intensity of MVPA
bouts(≥10consecutiveminutesspentinMVPA)[22].TheclassificationofMVPA
bouts was motivated by the PA recommendations of the Centers for Disease
Control and Prevention and the American College of Sports Medicine [23] and
theBritishAssociationofSportandExerciseSciences[24],whichcallforMVPA
to be accumulated in bouts of ≥10 min to achieve health benefits. Moreover,
these recommendations call for at least 150 min/week of bouted MVPA. We
investigated whether the participants in our study complied with these
recommendations,accountingforthenumberofvalidaccelerometrydays.
Skeletalmusclebiopsy
Biopsiesofthequadricepsmuscle(vastuslateralis)fromthedominantlegwere
obtained using the needle biopsy technique [25]. Muscle tissue was frozen in
melting isopentane precooled in liquid nitrogen and stored at −80°C for
histologicalpurposes.Anotherspecimenwassnap‐frozeninliquidnitrogenand
storedat−80°Cforgeneandproteinexpressionanalysesandforenzymeactivity
assays.
Fibersizeandcompositionanalysis
5 μm serial cryosections with fibers in transverse orientation were cut from
OCT‐embeddedmusclebiopsiesonacryostatmicrotome(LeicaCM1900,Meyer
Instruments, USA) at −20°C and mounted on SuperFrost microscope slides
(Menzel‐Gläser, Braunschweig, Germany) to be stored at −80°C until further
analyses. For immunohistochemistry, sections were incubated with primary
anti‐myosin heavy chain (MyHC) I (Developmental Studies Hybridoma Bank
[DSHB],UniversityofIowa,USA),anti‐MyHCIIa(DSHB,UniversityofIowa,USA)
and anti‐laminin (Sigma, Zwijndrecht, the Netherlands) followed by secondary
antibodies labeled with Alexa Fluor 555, Alexa Fluor 488 and Alexa Fluor 350
(Invitrogen,Breda,theNetherlands).Fibertypingwasaidedbymeansofmyosin
ATPase‐activity staining with acidic pre‐incubation at pH 4.40 [26].
Immunofluorescence‐stainedandATPase‐stainedsectionsweremicroscopically
photographed at 10× magnification. In a blinded fashion, fibers were classified
primarily based on immunofluorescence, with ATPase‐stained sections used to
confirmtypeI/IIaandtypeIIa/IIxhybridfibertypes.Fibercross‐sectionalarea
was measured with Lucia 4.82 software (Laboratory Imaging, Czech Republic)
basedonlamininstainingofthebasementmembrane[27].
46
CHAPTER3
RNAextractionandRT‐qPCRanalysis
10–30 mg muscle tissue was homogenized in Denaturation Solution (ToTALLY
RNA™ Kit; Ambion Ltd., Foster City, CA, USA) using a Polytron PT 1600 E
(Kinematica AG, Littau/Luzern, Germany) and RNA extracted according to the
supplier’s protocol, followed by genomic DNA removal and clean‐up with the
RNeasyMiniKitwithRNase‐freeDNase(Qiagen,Venlo,TheNetherlands).After
elution, RNA concentration was determined using a spectrophotometer
(NanoDrop ND‐1000, Isogen Lifescience, IJsselstein, The Netherlands) and
integrity verified for a selection of samples by gel electrophoresis and on a
Bioanalyzer (Agilent Technologies, Amstelveen, The Netherlands). 400 ng RNA
wasreversetranscribedtocDNAwithanchoredoligo(dT)primersaccordingto
the supplier’s protocol (Transcriptor First Strand cDNA Synthesis kit, Roche
Diagnostics,Woerden,TheNetherlands).RT‐qPCRprimersweredesignedbased
on Ensembl transcript sequences or selected from literature [28, 29] and
ordered from Sigma Genosys (Zwijndrecht, the Netherlands). The genes
encodingPGC‐1α,PGC‐1β,PGCrelatedcoactivator(PRC),PPAR‐α,Tfam,Nuclear
Respiratory Factor (NRF)‐1, NRF‐2α, Myosin Heavy Chain (MHC) I, muscle
phosphofructokinase (PFKM), MHCIIa and MYHCIIx were the target genes. RT‐
qPCRreactionscontainedSensiMixSYBRHi‐ROXKit(Quantace‐Bioline,London,
UK)with300nMprimersandwererunin384‐wellMicroAmpOptical384‐Well
Reaction Plate (Applied Biosystems, Nieuwerkerk a/d IJssel, The Netherlands)
ona7900HTFastReal‐TimePCRSystem(AppliedBiosystems).Standardcurves,
preparedfrompooledcDNA,andmeltcurveswereanalyzedtoverifyefficiency
and specificity of amplification. Twelve reference genes (ALAS1, ACTB, B2M,
PPIA,GAPDH,GUSB,HMBS,HPRT1,RPL13A,RPLP0,UBC,YWHAZ)weremeasured
and stability of expression was assessed by visual inspection of expression
differencesbetweenthestudygroupsandastabilityassessmentbygeNorm[29].
Eventually,ninereferencegenes(allexceptACTB,GAPDHandUBC)wereusedto
calculateageNormfactor,whichwasusedtonormalizeexpressionlevelsofthe
targetgenes.
ImmunoblottingofOXPHOSsubunitsandoxidativestressmarkers
10–20 mg tissue was crushed in liquid nitrogen and homogenized in 400 μl IP
lysisbuffer(50mMTris,150mMNaCl,10%glycerol,0.5%NonidetP40,1mM
EDTA, 1 mM Na3VO4, 5 mM NaF, 10 mM β‐glycerophosphate, 1 mM Na4O7P2, 1
mMdithiothreitol,10μg/mlleupeptin,1%aprotenin,1mMPMSF,pH7.4)witha
Polytron PT 1600 E (Kinematica). After homogenization, samples were
incubatedfor15minutesonarotatingwheelat4°Candspunfor30minutesat
maximum speed (20817×g) in a centrifuge cooled to 4°C. Supernatant was
aliquoted,snap‐frozenandstoredwithoutsamplebufferat−80°Cuntilanalysis.
47
Proteinconcentrationinlysateswasdeterminedusingthebicinchoninicacid
assay (Pierce, Thermo Fisher Scientific, Breda, The Netherlands). For western
blotanalysis,aliquotsweresupplementedwith4×samplebuffer(250mMTris‐
HCl pH 6.8, 8% (w/v) sodium dodecyl sulfate, 40% (v/v) glycerol, 0.4 M
dithiothreitol,0.02%(w/v)bromophenolblue)andkeptonice.Persample,5μg
unboiled protein was separated on gel (4‐12% Bis‐Tris XT gel, Criterion, Bio‐
Rad)withXTMOPSrunningbuffer(Bio‐Rad).Onesamplewasloadedonallgels
to facilitate gel–gel comparisons. Proteins were transferred to a 0.45 μm
nitrocellulosemembrane(Protran,SchleicherandSchuell,‘s‐Hertogenbosch,The
Netherlands) in transfer buffer (25 mM Tris, 192 mM glycine, 20% (v/v)
methanol) by electrophoresis. After transfer, membranes were blocked from
nonspecific protein binding with blocking solution, which contains 5% (w/v)
non‐fatdrymilk(Campina,Eindhoven,TheNetherlands)inTris‐bufferedsaline
with Tween20 (TBST; 25 mM Tris, 137 mM NaCl, 2.7 mM KCl, 0.05% (v/v)
Tween20,pH7.4),foronehouratroomtemperature,followedbyincubationin
primary antibody solution overnight at 4°C (mouse anti‐Total OXPHOS Rodent
WB Antibody Cocktail (MS604‐300, Abcam, Cambridge, UK) diluted 1:1000 in
blocking solution or rabbit anti‐GAPDH (#2118, Cell Signaling Technology,
Leiden, The Netherlands) diluted 1:5000 in TBST). Membranes were incubated
insecondaryantibodysolution(peroxidase‐labeledhorseanti‐mouseIgGorgoat
anti‐rabbit IgG (PI‐2000 and PI‐1000, respectively; Vector Laboratories,
Burlingame,CA,USA)diluted1:5000inblockingsolution)foronehouratroom
temperature before incubation with enhanced chemiluminescence substrate
(Pierce SuperSignal West PICO Chemiluminescent Substrate; Thermo Fisher
Scientific). Protein bands were detected using Super RX films (Fujifilm,
Düsseldorf, Germany)andscanned onaGS‐800 densitometer (Bio‐Rad). Bands
were quantified using Quantity One software (v4.6.2, Bio‐Rad), with GAPDH as
loading control (GAPDH levels were not different between COPD patients and
controls).
Protein content of oxidative stress markers were identified using specific
primary antibodies for protein carbonylation (anti‐2,4‐dinitrophenylhydrazone
[DNP]moietyantibody,Oxyblotkit,ChemiconInternationalInc.,Temecula,CA),
total protein nitration (anti‐3‐nitrotyrosine antibody, Invitrogen, Eugene,
Oregon),andtotalmalondialdehyde(MDA)‐proteinadducts(anti‐MDAantibody,
AcademyBio‐ Medical Company, Inc. Houston). For protein carbonylation,
carbonyl groups in the protein side chains were first derivatized to 2,4‐DNP
usingtheOxyblotkit(ChemiconInternationalInc.,Temecula,CA,USA)according
to the manufacturer’s instructions. Briefly, 15 g of protein were used per
derivatization reaction. Proteins were then denatured by addition of 12% SDS.
The samples were subsequently derivatized by adding 10 l of 1X 2,4‐DNP
solutionand incubated for 20 minutes. Finally, 7.5 l of neutralization solution
and2‐mercaptoethanolwereaddedtothesamplemixture.Immunoblottingwas
similar as described above with some minor differences: polyvinylidene
48
CHAPTER3
difluoride (PVDF) membranes were used, which were scanned with the
Molecular Imager Chemidoc XRS System (Bio–Rad Laboratories, Hercules, CA,
USA)andbandswerequantifiedusingthesoftwareImageLabversion2.0.1(Bio‐
Rad Laboratories). Values of total reactive carbonyl groups, total protein
tyrosine nitration, and total MDA‐protein adducts in a given sample were
calculatedbyadditionofopticaldensities(arbitraryunits)ofindividualprotein
bands in each case. Final optical densities obtained in each specific group of
subjects corresponded to the mean values of the different samples (lanes) of
each of theantigens studied. Actin (anti‐alpha‐sarcomeric actin antibody, clone
5C5,Sigma‐Aldrich,St.Louis,MO,USA)wasusedastheloadingcontrolforallthe
oxidativestressmarkers[12,30].
Enzymeactivityassays
15‐30mgtissuewascrushedinliquidnitrogenandhomogenizedin350μlSET
buffer (250 mM sucrose, 2 mM EDTA, 10 mM Tris, pH 7.4) with a Polytron PT
1600 E (Kinematica). After homogenization, samples were incubated for 10
minutes on a rotating wheel at 4°C and spun for 5 minutes at maximum speed
(20817×g)inacentrifugecooledto4°C.Analiquotofsupernatantwasstoredfor
proteindetermination.Totheremainingsupernatant,5×aqueousBSAsolution
wasaddedtoafinalBSAconcentrationof1%(v/v)andsampleswerealiquoted,
snap‐frozen and stored at −80°C until analysis. Protein concentration in SET
lysate aliquots without BSA was determined using the bicinchoninic acid assay
(Pierce). Citrate synthase (CS, EC 2.3.3.1), 3‐hydroxyacyl‐CoA dehydrogenase
(HADH,EC1.1.1.35)andphosphofructokinase(PFK,EC2.7.1.11)activitieswere
assayed spectrophotometrically (Multiskan Spectrum; Thermo Labsystems,
Breda, The Netherlands) as previously described [31‐33]. Absolute CS, HADH
andPFKactivitieswerenormalizedtototalprotein.
Statisticalanalysis
Differences between COPD patients and controls were tested using Student’s t‐
test, Mann‐Whitney‐U tests or χ2 tests as appropriate. Correlationswere tested
usingPearson’scorrelationcoefficientorSpearman’sρincaseofnon‐normally
distributeddata.AnalyseswereperformedusingPASWStatistics17.0(SPSSinc.
Chicago,IL,USA).AP‐value<.05wasdeclaredtobestatisticallysignificant.
Results
ThecontrolsubjectswerematchedtotheCOPDpatientsbasedonage,sexand
BMI(Table1).FEV1wasbetween30‐50%ofpredictedinn=13patients(45%),
between50‐70%ofpredictedinn=11patients(38%)and>70%ofpredictedin
n=5 patients (17%). VO2 peak and W peak were significantly decreased in the
49
COPDpatients(Table1).FEV1(%ofpredicted)wasstronglycorrelatedwithVO2
peak(%ofpredicted)inthetotalpopulation(r=0.83,P<.001),withintheCOPD
patients(r=0.78,P<.001),andwithinthecontrols(r=0.53,P=.044).
Table1.MaincharacteristicsofhealthycontrolsandCOPDpatients.
Controls(n=15)
COPD(n=29)
Demographics
Age,y
65(6)
65(6)
Sex,%men
60
55
Eversmoker,%
53
100a
BMI,kg/m2
24.9(3.3)
25.5(3.6)
Pulmonaryfunction
FEV1,%pred
113(15)
58(16)a
FVC,%pred
120(17)
104(22)b
FEV1/FVC,%
74(5)
45(11)a
DLCO,%pred
95(19)
53(18)a
Incrementalloadtest
PeakVO2,%pred
123(8)
72(4)a
PeakWR,%pred
133(7)
61(4)a
ΔBorgdyspneascore
3.4(0.8)
3.6(0.4)
ΔBorglegfatiguescore
2.7(0.6)
2.3(0.4)
ΔPaO2,kPa
1.04(0.53)
‐0.38(0.31)b
ΔLactate/WR(mmol/L/W)
0.039(0.003)
0.035(0.003)
Abbreviations:BMI,bodymassindex;DLCO,diffusioncapacityofthelungsforcarbonmonoxide;FEV1,
forcedexpiratoryvolumein1s;FVC,forcedvitalcapacity;WR,workrate.aP<.001,bP<.05.
Quadricepsmuscleenduranceandmarkersofoxidativephenotype
Quadriceps muscle endurance was significantly lower in the COPD patients
(Figure 1A). The mean decline in peak torque per repetition was ‐1.67±0.09
%N·m/rep in the COPD patients versus ‐1.18±0.10 %N·m/rep in the controls
(P=.002). Patients had a lower proportion of type I fibers and a higher
proportionoftypeIIa,typeIIa/IIxandIIxfiberscomparedtocontrols,indicative
of a fiber type I‐to‐II shift (Figures 1B‐D, and Figure 2). The mean mRNA
expressionlevelsofthekeyregulatorofoxidativemetabolismPGC‐1αandofthe
structuralmarkerofoxidativefibersMyHCIweresignificantlydecreasedinthe
COPD patients compared to controls, and reached borderline significance for
Tfam (Figure 1E). Other (co‐) transcription factors implicated in
50
CHAPTER3
Figure1.Quadricepsmuscleendurance,fibertypedistribution,geneexpression,
OXPHOSproteinexpressionandenzymeactivity.
A) Quadriceps endurance as assessed by isokinetic dynamometry. The P value denotes the statistical
significance of the difference in slopes between COPD patients and healthy controls. B) Quadriceps
musclefibertypedistributionasdeterminedbyfluorescentimmunohistochemistryandATPasestaining.
C) Boxplot of fiber type IIa/IIx proportion. Boxes represent 25th and 75th percentiles, and whiskers
represent10thand90thpercentiles.D)BoxplotoffibertypeIIxproportion.Boxesrepresent25thand75th
percentiles,andwhiskersrepresent10thand90thpercentiles.E)QuadricepsmusclemRNAexpressionof
peroxisome proliferator‐activated receptor (PPAR)‐γ coactivator‐1α (PGC‐1α), myosin heavy chain
51
(MyHC) I, IIa and IIx, mitochondrial transcription factor (Tfam), PPAR‐α, PGC‐1β, nuclear respiratory
factor (NRF)‐1, NRF‐2α, PGC related coactivator (PRC), and phosphofructokinase muscle (PFKM) as
determined by RT‐qPCR. F) Representative Western Blot of the five OXPHOS protein subunits and
GAPDHfromthreehealthycontrolsandthreeCOPDpatients.Allbandsarefromthesameblotandwere
reorganized for illustrative purposes as patients and controls were randomized over the blot and to
respect the ascending order of the five subunits. G) Levels of the five OXPHOS protein subunits as
determined by Western Blot. H) 3‐hydroxyacyl‐CoA dehydrogenase (HADH), citrate synthase (CS) and
phosphofructokinase (PFK) enzyme activities. I) Ratios between HADH/PFK and CS/PFK. *P<.05.
**P<.01.***P<.001.
drivingoxidative gene expression (PPAR‐α, PGC‐1β,PRC, NRF‐1, NRF‐2α) were
not differentially expressed between patients and controls. Gene expression
levels of the key glycolytic enzyme PFKM and of the structural markers of
glycolyticfibers,i.e.MyHCIIaandMyHCIIx,weresignificantlyhigherintheCOPD
patients(Figure1F).Congruently,meanOXPHOSsubunitproteinlevelswereall
lowerintheCOPDpatientscomparedtocontrols,reachingstatisticalsignificance
forthoseofcomplexesIVandV(Figures1F‐G).ThemeandifferencesinHADH,
CSandPFKenzymeactivitywerenotstatisticallysignificant(Figure1H),yetthe
ratioHADH/PFKwassignificantlylowerintheCOPDpatients.TheratioCS/PFK
reachedborderlinesignificance(Figure1I).
Figure 2. Peripheral muscle fiber type I‐to‐II shift, as an example of loss of
oxidativephenotypeinCOPD.
Representative myosin heavy chain ATPase stainings of quadriceps muscle biopsies from a healthy
person(A)andaCOPDpatient(FEV167%ofpredicted)(B)matchedforage,sexandbodymassindex
(25kg/m2).Blackfibers indicatetypeIfibersanddarkandlight grey fiberscorrespond withtypeIIX
andtypeIIAfibers,respectively.
Quadricepsstrength,musclemassandfibercross‐sectionalarea
FFMInorquadricepspeaktorqueweresignificantlydifferentbetweentheCOPD
patients and the controls (Figures 3A‐B). In line, the prevalence of muscle
wasting was not different between the COPD patients and the controls (20.7%
versus 13.3%, respectively, P=.70). Congruently, there were no differences in
musclefiberCSA(Figure3C).
52
CHAPTER3
Figure 3. Fat‐free mass index, quadriceps strength and fiber cross‐sectional
area.
A) Fat‐free mass index as determined by whole‐body dual‐energy X‐ray absorptiometry, P>.05. B)
Quadriceps peak torque as determined by isokinetic dynamometry, P>.05. C) Quadriceps muscle fiber
cross‐sectionalareaspresentedfortotalfibersandperfibertype,allP>.05.
Dailylivingphysicalactivitylevel
From all the subjects, 5.5±0.9 valid accelerometry days were available with a
mean of 14±1 hours of wear time per day. Total PA was significantly lower in
COPD patients compared with controls, and patients spent more time in
sedentary behavior and less time in MVPA (Table 2). Also, patients had fewer
MVPA bouts, which were shorter and less intense compared with the MVPA
boutsofcontrols(Table2).TotalPA(counts/min)waspositivelycorrelatedwith
FEV1 (% of predicted) in the entire study population (r=0.58, P<.001) and also
within the COPD patients (r=0.55, P=.002) but not in the controls (r=‐0.06,
P=.82). Within the COPD patients, FEV1 (% of predicted) was negatively
correlatedwithtimespentinsedentaryPA(r=‐0.54,P=.002)andpositivelywith
time spent in MVPA (r=0.61, P<.001), number of MVPA bouts per day (r=0.46,
P=.013),andwithtimespentinMVPAbouts(r=0.41, P=.039).Only31%ofthe
patientsfulfilledthecriteriaof150min/wkofboutedMVPAtimeversus80%of
thecontrols(P=.004).Withinthepatients,thosewhofulfilledthesecriteriahada
significantly higher FEV1 compared to the patients who did not (67±5 versus
54±3%ofpredicted,P=.049).
AssociationswithquadricepsoxpheninCOPDpatients
In the COPD patients, quadriceps endurance was significantly positively
correlated with myosin heavy chain I gene expression and PGC‐1α gene
expression,andshowedatrendtowardsapositivecorrelationwithtypeIfiber
proportion (Figures 4A‐C). In line, type IIx fiber proportion was negatively
correlatedwithquadricepsendurance(ρ=‐0.43,P=.038).Moreover,typeIfiber
53
Table2.AccelerometrydatafromCOPDpatientsandcontrols.
Controls(n=15) COPD(n=29)
Totalactivity,counts/min
TimespentinsedentaryPA,%ofwear time
TimespentinlifestylePA,%ofweartime
TimespentinMVPA,%ofweartime
TotalnumberofMVPAboutsd
NumberofMVPAbouts perdayd
TimespentinMVPAbouts,mind
MeanMVPAboutduration,min
MeanintensityofMVPAbouts,countse
CompliancewithMVPArecommendation,%f
360(134)
60.5(6.6)
25.5(4.6)
13.8(5.3)
12(8‐18)
2.0(1.3‐2.0)
267(108‐494)
22.1(9.4)
2700(605)
80
220(101)a
67.4(9.4)b
24.3(6.5)
8.6(5.3)b
5(3‐11)b
1.0(0.6‐2.2)c
68(38‐204)b
14.7(3.7)b
1913(498)a
3b
Abbreviations: MVPA, moderate‐to‐vigorous physical activity; PA, physical activity. aP<.001. bP<.01.
cP<.05. dMedian (IQR). eWeighted for individual MVPA bout duration. fAt least 150 min/wk of bouted
MVPAtime.
proportion was negatively correlated with the increase in lactate relative to
workrate(Figure4D).
No significant correlations were found between markers of oxphen and
quadriceps peak torque, pulmonary function, exercise capacity, total PA, time
spentinsedentaryPAorMVPA,oranyoftheMVPAbout‐relatedvariables(data
notshown).
Skeletalmuscleoxidativestress
Thelevelsofproteincarbonylation,tyrosinenitrationandMDA‐proteinadducts
werenotdifferentbetweenpatientsandcontrols(Figure5).Amongthepatients,
we found inverse correlations between protein carbonylation and type I fiber
proportion, myosin heavy chain I gene expression and quadriceps endurance,
andapositivecorrelationbetweenproteincarbonylationlevelandmyosinheavy
chain IIa expression (Figure 6). In line, protein carbonylation was positively
correlatedwithtypeIIxfiberproportion(ρ=0.43,P=.023).Additionally,tyrosine
nitrationwasinverselycorrelatedwithquadricepsendurance(r=‐0.42,P=.040),
and MDA‐protein adductswere inversely associatedwith myosinheavy chain I
gene expression (r=‐0.38, P=.044). Markers of oxidative stress were not
correlated with FFMI or fiber CSA among patients, and no correlations were
found between oxidative stress markers and markers of oxphen in the healthy
subjects(datanotshown).
54
CHAPTER3
Figure4.CorrelationsamongCOPDpatientsbetweenmarkersofskeletalmuscle
oxidative phenotype and quadriceps endurance and the rise in blood lactate
relativetoworkrateduringanincrementalcyclingtest.
A)CorrelationbetweentheproportiontypeIfibersandquadricepsendurance.B)Correlationbetween
myosin heavy chain I mRNA expression and quadriceps endurance. C) Correlation between PGC‐1α
mRNA expression and quadriceps endurance. D) Correlation between the proportion type I fibers and
theriseinbloodlactaterelativetoworkrateduringanincrementalcyclingtest.
Discussion
Whereas previous studies have clearly shown skeletal muscle dysfunction in
advancedCOPD,themainnoveltyofthecurrentstudyisthatweidentifiedaloss
of quadriceps oxphen and decreased quadriceps endurance in patients with
mild‐to‐moderate COPD. Interestingly, these abnormalities existed even in the
absence of significant muscle wasting. Also we found consistent inverse
correlations between markers of oxphen and oxidative stress among the
patients.Itshouldbepointedoutthatourstudyisofcross‐sectionalnature,and
does not allow for longitudinal inferences. However, linking the various cross‐
55
Figure5.Levelsofskeletalmuscleoxidativestress.
Thelevelsofproteincarbonylation,tyrosinenitrationandMDA‐protein
adducts were not different between COPD patients and controls, all
P>.05.
sectionalstudiesacrossCOPDseveritystages,theavailabledatasuggestthatthe
loss of skeletal muscle endurance and skeletal muscle oxphen occur already in
patientswithmild‐to‐moderateCOPDandprogresswithdeclininglungfunction,
andsuggestpotentialinvolvementofoxidativestress.
Weassessedalargepanelofmarkersofskeletalmuscleoxphen,andfound
that many were significantly decreased in our mild‐to‐moderate COPD patients
comparedtowell‐matchedcontrols.Morespecifically,thelossofoxpheninour
COPD patients was characterized by a decreased proportion of type I fibers,
decreasedgeneexpressionofPGC‐1αandMyHCI,decreasedproteinexpression
of subunits of OXPHOS complexes IV and V, and decreased HADH/PFK enzyme
activity.ThemeandifferencesinproportionsoftypeI,IIaandIIxfibersbetween
our mild‐to‐moderate COPD patients and matched healthy controls were 20%,
15% and 4%, respectively. For comparison, pooled analyses from a meta‐
analysisinadvancedCOPDpatientsshowedthesedifferencestobe22%,7%and
13%, respectively [7]. These combined data suggest that a decrease in slow‐
oxidativetypeIfiberproportionoccursalreadyinearlyCOPD,andthatafurther
shift from fast‐oxidative type IIa to fast‐glycolytic type IIx fibers continues in
advanced COPD towards even more dependence on glycolytic metabolism. It
should be acknowledged that decreased expression of subunits of the key
respiratorychaincomplexesdoesnotnecessarilyindicatedecreasedcytochrome
coxidase(COX)activityoroveralllowermitochondrialrespiration.Severalother
studies have assessed COX activity in muscle biopsies of COPD patients, and
showed contrasting results. For example, reduced skeletal muscle COX activity
has been reported in moderate COPD patients by some authors [34‐
56
CHAPTER3
Figure6.CorrelationsamongCOPDpatientsbetweenmarkersofskeletalmuscle
oxidative phenotype, quadriceps endurance and the level of protein
carbonylation.
Correlations between the level of protein carbonylation and proportion type I fibers (A), quadriceps
endurance(B),myosinheavychainmRNAexpression(C),andmyosinheavychainIIamRNAexpression(D).
36], whereas others reported even higher COX activity in moderate‐to‐severe
COPD patients [37, 38]. Between‐study differences remain unexplained.
Therefore, these data provide further indication for studying mitochondrial
respiration in more detail in different COPD disease stages, including the early
stages.
Wefoundthatmarkersofoxidativephenotypeweresignificantlycorrelated
with quadriceps endurance, implying that the magnitude of loss of oxidative
phenotype reached the level of clinical importance in terms of quadriceps
endurance. Similar findings were previously reported in a population of
advancedCOPDpatientswithsignificantmusclewasting[4].Ourdatashowthat
these abnormalities are already present in early COPD even in the absence of
muscle wasting. In a recent study by Saey et al. [39] Δlactate/WR during a
constantworkratetestwassignificantlyhigherinmuscle‐wastedCOPDpatients
(mean FEV1 45% of predicted) compared with controls. This coincided with a
decreased proportion of type I fibers and increased type IIa fibers and a wide
57
array of increased glycolytic markers. The authors argued that increased
glycolysis underlies the increased Δlactate/WR. Although this seems plausible,
nocorrelationsbetweenΔlactate/WRandmetabolicmarkerswerepresentedby
Saey et al. [39]. We however did find that the proportion of type I fibers was
associated with increased Δlactate/WRduring an incremental load test in non‐
musclewastedCOPDpatients.
PGC‐1α is considered the major regulator of skeletal muscle oxidative
metabolism. Indeed, its expression is known to be significantly induced upon
aerobic exercise, which subsequently orchestratesoxidativegene expression in
ordertopreparethemuscleforanextboutofexercise[40].Puente‐Maestuetal.
[41] have recently shown that moderate‐intensity exercise increased skeletal
musclePGC‐1αgeneexpressioninCOPDpatients(FEV150%ofpredicted)with
normalFFMI.AswefoundthatPGC‐1αgeneexpressionwasdecreasedby40%
inourpatients,thiswouldsuggestthatthemuscleoxidativemachineryfails,at
leastinpart,attheregulatorylevel.PGC‐1α’sdownstreamtargetTfamtendedto
be decreased expressed as well. Expression levels of other (co‐) transcription
factorsimplicatedintheregulationofTfamexpression(PGC‐1β,PRC,NRF‐1,and
NRF‐2α)werenotdifferentiallyexpressedbetweenpatientsandcontrols.
Physical inactivity, or deconditioning, has been proposed to contribute to
skeletalmuscledysfunctioninCOPD[42].Weusedaccelerometrytoassessboth
quantitative and qualitative daily living physical activity, and studied the
relationswithmuscleoxphen.Wedidnotspecificallyselectthehealthysubjects
based on sedentary behavior, so the finding that our COPD patients were on
average 39% less physically active compared with the controls was not
surprising.Inaddition,wefoundthatourpatientsengagedin lessMVPAbouts
andtheseboutswereshorterandlessintensecomparedtothoseofourcontrols.
The level and intensity of physical activity was positively correlated with FEV1
within the COPD patients, which is consistent with previous reports [43‐45].
Interestingly, however, we could not identify any relation between markers of
oxphenandtotalphysicalactivitylevelnoritsintensityinourpatients.Although
we did not select our healthy controls on sedentary behavior, our data do not
supportthepresumptionthatphysicalinactivityistheprincipledeterminantof
impairedskeletalmuscleoxpheninmild‐to‐moderateCOPD.Still,thefactthatno
association between loss of oxphen and physical inactivity was found does not
exclude the possibility that physical activity plays a role in the multifactorial
pathwaysinvolvedinlossofoxphen.
Whereas basal levels of skeletal muscle oxidative stress were comparable
between our patients and controls, we found that markers of muscle oxidative
stressweresignificantlyassociatedwiththelossinmuscleoxphenamongCOPD
patients, but were not related to muscle mass. Oxidative stress has been
suggested to render the muscle susceptible to atrophy as damaged (oxidized)
proteins are prone to be degraded. However, this concept has recently been
challengedasnodifferencesinbasalskeletalmuscleoxidativestresslevelswere
58
CHAPTER3
foundbetweenadvancedCOPDpatientswithandwithoutmusclewasting[12].
Enhanced mitochondrial production of reactive oxygen species (ROS) has been
observedinpatientswhichmoderate‐to‐severeCOPD,whichhasbeenproposed
to be linked to the proportional increase in type II fibers [46]. Recently, it has
also been demonstrated that OXPHOS complex III is the main site of ROS
production within peripheral muscles of COPD patients, who exhibited similar
clinicalfeaturesasthoserecruitedinthecurrentinvestigation[9].Furthermore,
compared to type I fiber mitochondria, type II fiber mitochondria exhibit
enhanced ROS production, as was shown in rats [47]. Taken together, the
associationsencounteredbetweenproteinoxidationlevelsandtheslow‐to‐fast
phenotypeshiftamongmild‐to‐moderateCOPDpatientsmaybepartlyexplained
bymitochondrialdysfunctionandenhancedROSproduction.
Inperipheralskeletalmuscleofemphysematoushamsters,decreasedcitrate
synthase activity and increased lipid peroxidation have been reported, which
were not related to decreased body weight, muscle mass or physical activity
level[10,11].Inline,inaratmodelofemphysema,Zhangetal.[48]reportedno
change in muscle mass but showed increased skeletal muscle lipofuscin
inclusions(amarkerofoxidativestress)whichwerecorrelatedwithdecreased
muscle endurance. Collectively, the available data suggest a relation between
oxidative stress and the loss of oxidative capacity in COPD‐related muscle
dysfunction and it can be speculated that loss of oxphen, through enhanced
oxidativestress,mayeventuallyaugmentmusclewastingaswell.
In conclusion, this study shows evidence for a loss of muscle oxphen and
decreased quadriceps endurance in patients with mild‐to‐moderate COPD
withoutovertlossofmusclemass.Ourresultsindicatethattimelyintervention
strategies aimed at improving muscle oxphen in early COPD may improve or
preventafurtherlossinquadricepsendurance.
59
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
60
Skeletalmuscledysfunctioninchronicobstructivepulmonarydisease.AstatementoftheAmericanThoracic
SocietyandEuropeanRespiratorySociety.AmJRespirCritCareMed.1999;159(4Pt2):S1‐40.
Bernard S, LeBlanc P, Whittom F, Carrier G, Jobin J, Belleau R, Maltais F. Peripheral muscle weakness in
patientswithchronicobstructivepulmonarydisease.AmJRespirCritCareMed.1998;158(2):629‐34.
vanWeteringCR,HoogendoornM,MolSJ,Rutten‐vanMolkenMP,ScholsAM.Short‐andlong‐termefficacyof
a community‐based COPD management programme in less advanced COPD: a randomised controlled trial.
Thorax.2010;65(1):7‐13.
ScholsAM,GoskerHR.Thepathophysiologyofcachexiainchronicobstructivepulmonarydisease.CurrOpin
SupportPalliatCare.2009;3(4):282‐7.
2007;62(11):944‐9.
Remels AH, Schrauwen P, Broekhuizen R, Willems J, Kersten S, Gosker HR, Schols AM. Peroxisome
proliferator‐activated receptor expression is reduced in skeletal muscle in COPD. Eur Respir J.
2007;30(2):245‐52.
Puente‐MaestuL,TejedorA,LazaroA,deMiguelJ,Alvarez‐SalaL,Gonzalez‐AragonesesF,SimonC,AgustiA.
SiteofMitochondrialROSProductioninSkeletalMuscleofCOPDanditsRelationshipwithExerciseOxidative
Stress.AmJRespirCellMolBiol.2012;47(3):358‐62.
MattsonJP,PooleDC.Pulmonaryemphysemadecreaseshamsterskeletalmuscleoxidativeenzymecapacity.J
ApplPhysiol.1998;85(1):210‐4.
Mattson JP, Sun J, Murray DM, Poole DC. Lipid peroxidation in the skeletal muscle of hamsters with
emphysema.Pathophysiology.2002;8(3):215‐21.
Fermoselle C, Rabinovich R, Ausin P, Puig‐Vilanova E, Coronell C, Sanchez F, Roca J, Gea J, Barreiro E. Does
oxidativestressmodulatelimbmuscleatrophyinseverecopdpatients?EurRespirJ.2012Mar9[Epubahead
ofprint].
BarreiroE,GeaJ,CorominasJM,HussainSN.Nitricoxidesynthasesandproteinoxidationinthequadriceps
femorisofpatientswithchronicobstructivepulmonarydisease.AmJRespirCellMolBiol.2003;29(6):771‐8.
QuanjerPH,TammelingGJ,CotesJE,PedersenOF,PeslinR,YernaultJC.Lungvolumesandforcedventilatory
flows.ReportWorkingPartyStandardizationofLungFunctionTests,EuropeanCommunityforSteelandCoal.
OfficialStatementoftheEuropeanRespiratorySociety.EurRespirJSuppl.1993;16:5‐40.
FranssenFM,SauerweinHP,AckermansMT,RuttenEP,WoutersEF,ScholsAM.Increasedpostabsorptiveand
exercise‐induced whole‐body glucose production in patients with chronic obstructive pulmonary disease.
Metabolism.2011;60(7):957‐64.
van den Borst B, Gosker HR, Wesseling G, de Jager W, Hellwig VA, Snepvangers FJ, Schols AM. Low‐grade
adiposetissueinflammationinpatientswithmild‐to‐moderatechronicobstructivepulmonarydisease.AmJ
ClinNutr.2011;94(6):1504‐12.
ScholsAM,SoetersPB,DingemansAM,MostertR,FrantzenPJ,WoutersEF.Prevalenceandcharacteristicsof
nutritionaldepletion in patientswith stable COPD eligiblefor pulmonary rehabilitation. Am Rev Respir Dis.
1993;147(5):1151‐6.
FranssenFM,BroekhuizenR,JanssenPP,WoutersEF,ScholsAM.LimbmuscledysfunctioninCOPD:effectsof
musclewastingandexercisetraining.MedSciSportsExerc.2005;37(1):2‐9.
WagenmakersR,vandenAkker‐ScheekI,GroothoffJW,ZijlstraW,BulstraSK,KootstraJW,Wendel‐VosGC,
vanRaaijJJ,StevensM.Reliabilityandvalidityoftheshortquestionnairetoassesshealth‐enhancingphysical
activity(SQUASH)inpatientsaftertotalhiparthroplasty.BMCMusculoskeletDisord.2008;9:141.
Troiano RP, Berrigan D, Dodd KW, Masse LC, Tilert T, McDowell M. Physical activity in the United States
measuredbyaccelerometer.MedSciSportsExerc.2008;40(1):181‐8.
MatthewCE.Calibrationofaccelerometeroutputforadults.MedSciSportsExerc.2005;37(11Suppl):S512‐22.
Masse LC, Fuemmeler BF, Anderson CB, Matthews CE, Trost SG, Catellier DJ, Treuth M. Accelerometer data
reduction: a comparison of four reduction algorithms on select outcome variables. Med Sci Sports Exerc.
2005;37(11Suppl):S544‐54.
PateRR,PrattM,BlairSN,HaskellWL,MaceraCA,BouchardC,BuchnerD,EttingerW,HeathGW,KingAC,et
O'DonovanG,BlazevichAJ,BorehamC,CooperAR,CrankH,EkelundU,FoxKR,GatelyP,Giles‐CortiB,GillJM,
etal.TheABCofPhysicalActivityforHealth:aconsensusstatementfromtheBritishAssociationofSportand
ExerciseSciences.JSportsSci.2010;28(6):573‐91.
Bergstrom L. Muscle electrolytes in man. Determination by neutron activation analysis on needle biopsy
specimens.Astudyonnormalsubjects,kidneypatients,andpatientswithchronicdiarrhea.ScandJClinLab
Invest.1962;68:1–110.
OgilvieRW,FeebackDL.Ametachromaticdye‐ATPasemethodforthesimultaneousidentificationofskeletal
musclefibertypesI,IIA,IIBandIIC.StainTechnol.1990;65(5):231‐41.
CHAPTER3
27. Verdijk LB, Koopman R, Schaart G, Meijer K, Savelberg HH, van Loon LJ. Satellite cell content is specifically
reducedintypeIIskeletalmusclefibersintheelderly.AmJPhysiolEndocrinolMetab.2007;292(1):E151‐7.
28. AllenD,WintersE,KennaPF,HumphriesP,FarrarGJ.Referencegeneselectionforreal‐timertPCRinhuman
epidermalkeratinocytes.JDermatolSci.2008;49(3):217‐25.
29. VandesompeleJ,DePreterK,PattynF,PoppeB,VanRoyN,DePaepeA,SpelemanF.Accuratenormalization
ofreal‐timequantitativeRT‐PCRdatabygeometricaveragingofmultipleinternalcontrolgenes.GenomeBiol.
2002;3(7):RESEARCH0034.
30. RodriguezDA,KalkoS,Puig‐VilanovaE,Perez‐OlabarriaM,FalcianiF,GeaJ,CascanteM,BarreiroE,RocaJ.
Muscle and blood redox status after exercise training in severe COPD patients. Free Radic Biol Med.
2012;52(1):88‐94.
31. BergmeyerHU,GawehnK,GrasslM.3‐Hydroxyacyl‐CoAdehydrogenase.In:BergmeyerHU,editor.Methodsof
enzymaticanalysis.Weinheim;Germany:VerlagChemieGmbH;1974.p.474.
32. Ling KH, Paetkau V, Marcus F, Lardy HA. Phosphofructokinase: I. Skeletal Muscle. Methods in Enzymology:
AcademicPress;1966.p.425‐9.
33. ShepherdD,GarlandPB.Citratesynthasefromratliver.MethodsinEnzymology:AcademicPress;1969.p.11‐6.
34. GoskerHR,vanMamerenH,vanDijkPJ,EngelenMPKJ,vanderVusseGJ,WoutersEFM,ScholsAMWJ.Skeletal
musclefibretypeshiftingandmetabolicprofileinpatientswithCOPD.EurRespirJ.2002;19:617‐26.
35. NaimiAI,BourbeauJ,PerraultH,BarilJ,Wright‐ParadisC,RossiA,TaivassaloT,SheelAW,RabolR,DelaF,et
al.AlteredmitochondrialregulationinquadricepsmusclesofpatientswithCOPD.ClinPhysiolFunctImaging.
2011;31(2):124‐31.
36. GreenHJ,BombardierE,BurnettM,IqbalS,D'ArsignyCL,O'DonnellDE,OuyangJ,WebbKA.Organizationof
metabolicpathwaysinvastuslateralisofpatientswithchronicobstructivepulmonarydisease.AmJPhysiol
RegulIntegrCompPhysiol.2008;295(3):R935‐41.
37. Sauleda J, Garcia‐Palmer F, Wiesner RJ,Tarraga S, Harting I, Tomas P, Gomez C, Saus C, Palou A, Agusti AG.
Cytochrome oxidase activity and mitochondrial gene expression in skeletal muscle of patients with chronic
38. Puente‐MaestuL,Perez‐ParraJ,GodoyR,MorenoN,TejedorA,Gonzalez‐AragonesesF,BravoJL,AlvarezFV,
CamanoS,AgustiA.AbnormalmitochondrialfunctioninlocomotorandrespiratorymusclesofCOPDpatients.
EurRespirJ.2009;33(5):1045‐52.
39. Saey D, Lemire BB, Gagnon P, Bombardier E, Tupling AR, Debigare R, Cote CH, Maltais F. Quadriceps
metabolism during constant workrate cycling exercise in chronic obstructive pulmonary disease. J Appl
Physiol.2011;110(1):116‐24.
40. PilegaardH,SaltinB,NeuferPD.ExerciseinducestransienttranscriptionalactivationofthePGC‐1alphagene
inhumanskeletalmuscle.JPhysiol.2003;546(Pt3):851‐8.
41. Puente‐Maestu L, Lazaro A, Tejedor A, Camano S, Fuentes M, Cuervo M, Navarro BO, Agusti A. Effects of
exerciseonmitochondrialDNAcontentinskeletalmuscleofpatientswithCOPD.Thorax.2011;66(2):121‐7.
42. Wagner PD. Skeletal muscles in chronic obstructive pulmonary disease: deconditioning, or myopathy?
Respirology.2006;11(6):681‐6.
43. Steele BG, Holt L, Belza B, Ferris S, Lakshminaryan S, Buchner DM. Quantitating physical activity in COPD
usingatriaxialaccelerometer.Chest.2000;117(5):1359‐67.
44. Jehn M, Schmidt‐Trucksass A, Meyer A, Schindler C, Tamm M, Stolz D. Association of daily physical activity
volumeandintensitywithCOPDseverity.RespirMed.2011;105(12):1846‐52.
45. Garcia‐RioF,LoresV,MedianoO,RojoB,HernanzA,Lopez‐CollazoE,Alvarez‐SalaR.Dailyphysicalactivityin
patientswithchronicobstructivepulmonarydiseaseismainlyassociatedwithdynamichyperinflation.AmJ
RespirCritCareMed.2009;180(6):506‐12.
46. Picard M, Godin R, Sinnreich M, Baril J, Bourbeau J, Perrault H, Taivassalo T, Burelle Y. The mitochondrial
phenotypeofperipheralmuscleinchronicobstructivepulmonarydisease:disuseordysfunction?AmJRespir
CritCareMed.2008;178(10):1040‐7.
47. AndersonEJ,NeuferPD.TypeIIskeletalmyofiberspossessuniquepropertiesthatpotentiatemitochondrial
H(2)O(2)generation.AmJPhysiolCellPhysiol.2006;290(3):C844‐51.
48. ZhangXL,PangBS,HouXL,WangJ,WangC.Oxidativestressandperipheralskeletalmuscledysfunctionin
ratswithemphysema.ChinMedJ(Engl).2010;123(1):40‐4.
61
CHAPTER4
CHAPTER4
Low‐grade adipose tissue inflammation
inpatientswithmild‐to‐moderateCOPD
Bram van den Borst, Harry R. Gosker, Geertjan Wesseling, Wilco de
Jager, Valéry A.C.V. Hellwig, Frank J. Snepvangers, Annemie M.W.J.
Schols
63
ADIPOSETISSUEINFLAMMATIONINCOPD
Abstract
Background: Low‐grade systemic inflammation is common in chronic
obstructivepulmonarydisease(COPD),butitssourceremainsunclear.Adipose
tissue is a potent producer of inflammatory mediators and may contribute to
systemicinflammationinCOPD,possiblyviahypoxia.
Objective: To study the influence of COPD and exercise‐induced oxygen
desaturation on adipose tissue inflammation (ATI) and its contribution to
systemicinflammation.
Design:Subcutaneousadiposetissuebiopsieswereinvestigatedof28clinically
stable COPD patients (FEV1 58±16 % of predicted, body mass index [BMI]
24.9±2.9kg/m2)and15matchedhealthycontrols.Fatmasswasmeasuredwith
dual‐energy X‐ray absorptiometry. Patients were pre‐stratified by oxygen
desaturation assessed by incremental cycle ergometry. Adipocyte size and
adipose tissue expression of 19 inflammatory and hypoxia‐related genes were
measured and adipose tissue macrophages (ATMs) were histologically
quantified. Systemic inflammatory markers included CRP and a panel of 20
adipokines.
Results:COPDpatientshadcomparablefatmassbuthigherCRPandHOMA‐IR
indexcomparedtocontrols.COPDpatientsandcontrolshadcomparableadipose
tissuegene expression, adipocytesize, ATM infiltration and systemicadipokine
levels.DesaturatingCOPDpatientshadnodifferentATIstatuscomparedtonon‐
desaturating patients. COPD patients with high CRP had significantly greater
ATMinfiltrationcomparedtopatientswithlowCRP,whichwasindependentof
BMIandfatmass.
Conclusions: We identified a possible role of ATMs in the low‐grade systemic
inflammation in COPD. ATI and systemic adipokine levels were comparable to
controlsandpartlyrelatedtofatmass.Oxygendesaturationwasnotassociated
withATI.
64
CHAPTER4
Introduction
Chronic obstructive pulmonary disease (COPD) is characterized by low‐grade
systemic inflammation which has been associated with insulin resistance (IR)
[1]. Classically, systemically elevated cytokine levels are hypothesized to result
from a ‘spill‐over’ of the pulmonary compartment. However, there is no
convincing evidence to support this theory in clinically stable disease.
Alternatively, adipose tissue has been proposed to contribute to systemic
inflammationinobeseCOPDpatients[2].Inadditiontostoringenergy,adipose
tissueisanactiveproducerofmediatorsinvolvedininflammation;theso‐called
adipokinesofwhichleptinandadiponectinareexamples[2].
Several mechanisms have been proposed to induce or relate to adipose
tissueinflammation(ATI),thatwehypothesizedcouldbeinvolvedinCOPDand
resultinenhancedsystemicinflammation.First,largeadipocytesproducemore
pro‐inflammatory adipokines than small adipocytes [3], indicative of adipocyte
size as a marker of ATI. Most patients with COPD receive chronic β2‐agonists
that could contribute to the blunted β‐adrenoreceptor mediated lipolysis and
thermogenesis resulting in maintenance or relative expansion of fat tissue in
COPD[4].Second,amajorreductionofoxygensupplymayleadtoATIandmay
potentially be aggravated via a disbalance between adipocyte size and (neo‐)
vascularisation leading to local hypoxic areas and/or intra‐adipocytic hypoxia
[5].Itisunknown,however,whethermildhypoxaemiaaspresentinearlystages
of COPD or intermittent desaturations in some COPD phenotypes could also
triggerATIinCOPD.Chronichypoxaemiacanoccurduringadvanceddiseasebut
recurrent desaturations also occur in moderate disease during daily physical
activities [6]. Third, IR has been linked to both systemic inflammation [7] and
ATI [8]. Finally, although adipose tissue macrophages (ATMs) have been put
forwardastheorchestratorsofATI[9]theyhaveneverbeenquantifiedinCOPD.
Recently, Tkacova et al. [10] showed higher pro‐inflammatory gene
expression in adipose tissue of underweight GOLD IV patients compared to
overweightGOLDI‐IIIpatients.Interestingly,groupswerealsodifferentinPaO2
(7.7vs9.4kPa,respectively)suggestingthathypoxaemiamayplayaroleinATI.
Moreover,thisstudysuggeststhatATImaynotberestrictedtotheobesestate.
Subsequently, the same authors showed that with increasing body mass index
(BMI) (range 18‐36 kg/m2), a higher prevalence of IR was present in COPD
patients which was associated with higher adipose tissue expression of pro‐
inflammatory genes [11]. BMI and fat mass are, however, known correlates of
ATI[8,12,13].Thus,toanswerthequestionifCOPDperseischaracterizedby
ATI and to study its relation with systemic inflammation, we studied COPD
patients and BMI‐ and fat mass‐matched healthy controls. Additionally, we
investigated a potential role of exercise‐induced desaturation in COPD on ATI
andsystemicinflammation.
65
Methods
Adetailedmethodologycanbefoundinthesupplementalmaterial.
Subjects,pulmonaryfunctionandincrementalexercisetest
This study included 28 clinically stable COPD patients without overt co‐
morbidity and 15 healthy controls. Pulmonary medication is listed in
Supplemental Table 1. Pulmonary function testing included forced spirometry
andsinglebreathdiffusioncapacitymeasurement(Masterlab,Jaeger,Würzburg,
Germany).Allsubjectsperformedanincrementalcycleergometrytestaccording
to international standards to determine exercise‐induced desaturation defined
as a fall in oxygen saturation (SaO2, by pulse oximetry) of ≥4% [14]. Twelve
COPDpatientsdesaturated(COPDD,mean±SDΔSaO2‐8.5±2.2%)and16didnot
(COPDND,ΔSaO2‐2.1±1.7%).Arterialbloodgasanalysis(Bloodgasanalyzer865,
ChironDiagnostics,Emeryville,CA)wasperformedatrestandatVO2maxfroma
radialarterypuncture.Writteninformedconsentwasobtainedfromallsubjects
andtheethicalreviewboardoftheMaastrichtUMC+approvedthestudy(MEC
08‐2‐059).Thisstudyisregisteredonwww.trialregister.nl(NTR1402).
Anthropometricsandbodycomposition
Height was measured using a wall‐mounted stadiometer. Body weight was
assessed to the nearest 0.1 kg using a standard balance beam scale. BMI was
calculated as weight/height2 (kg/m2). Whole‐body dual energy X‐ray
absorptiometry (DXA, Hologic Discovery, Hologic, or Lunar Prodigy, GE
Healthcare)wasappliedtoretrievewholebodyfatmass(FM)andfat‐freemass
(FFM). Comparison of Lunar and Hologic retrieved measures was enabled
through applying cross‐calibration equations (kindly provided by Dr. Shepherd
at the University of California, San Francisco, CA, USA, unpublished). Fat‐free
massindex(FFMI)andfatmassindex(FMI)werecalculatedasFFM/height2and
FM/height2,respectively.
Adiposetissuebiopsy
Subcutaneous adipose tissue (SAT) was obtained para‐umbilically after an
overnight fast through needle biopsy. One specimen was snap‐frozen in liquid
nitrogen and stored at ‐80°C until further analysis and another specimen was
processedin4%formalinforparaffinembedding.
66
CHAPTER4
Real‐timequantitativepolymerasechainreaction(RT‐qPCR)
Total RNA from SAT was extracted with the RNeasy® Lipid Tissue Mini Kit
(Qiagen, Venlo, The Netherlands) and reverse‐transcribed into cDNA using the
TranscriptorcDNAsynthesiskit(RocheAppliedSciences,Mannheim,Germany).
cDNAwasamplifiedusingSYBRGreenFluoresceinMix(Westburg,Leusden,The
Netherlands) on a quantitative PCR machine (Bio‐Rad laboratories B.V.,
Veenendaal, The Netherlands). We measured a panel of genes consisting of
adipokines (leptin, adiponectin, plasminogen activator inhibitor‐1 [PAI‐1],
chemerin, osteoprotegerin, adrenomedullin and visfatin), markers of hypoxia
and (neo)vascularisation (glucose transporter‐1 [GLUT1], hypoxia inducible
factor‐1α [HIF‐1α], vascular endothelial growth factor‐A [VEGF‐A] and CD34),
markers of macrophages (CD68, CD163, CD206, CD11b and monocyte
chemotactic protein‐1 [MCP‐1]), markers of inflammation (IL‐8 and IL‐10) and
peroxisomeproliferator‐activatedreceptor‐γ(PPAR‐γ)asamarkerofadipocyte
differentiation. GeNorm software (Primerdesign, Southampton, USA) was
applied to calculate a normalization factor based on the expression levels of
threehousekeepinggenes[15].PrimersequencesareprovidedinSupplemental
Table2.Geneexpressionwasquantifiedandexpressedasarbitraryunits(AU).
Adipocytesize
Adipocytesizewasdeterminedonhaematoxylin‐eosinstainedparaffinsections
byquantifyingtheactualareaof≥400uniqueadipocytesperindividualthrough
computerimageanalysissoftware(LuciaGF,Version4.81).
Quantificationofadiposetissuemacrophages(ATMs)
ATMs were quantified on adipose tissue paraffin sections that were incubated
with anti‐CD68, visualized by the EnVision™ FLEX/FLEX+ System
(DakoCytomation, Glostrup, Denmark) and counter‐stained with haematoxylin.
Through direct microscopy at 400x magnification, two blinded raters
systematicallyquantifiedthenumberofsolitaryATMsandcrown‐likestructures
(CLS)(Figure1)andtheirdensitieswereexpressedasnumberofsolitaryATMs
andCLSpermm2ofanalyzedtissue.
Cardiometabolicandsystemicinflammatoryprofile
Fasting plasma concentrations of 19 adipocytokines were analyzed with a
multipleximmunoassayasdescribedindetailelsewhere[16].Thelowerlimitsof
detection(LLOD)arepresentedinTableE3.FurthermeasurementsincludedC‐
reactive protein (CRP), N‐terminal pro‐Brain Natriuretic Peptide (NT‐proBNP),
glucoseandinsulin.Seeonlinesupplementformoredetails.Homeostaticmodel
67
assessment (HOMA) was applied to estimate IR by HOMA‐IR=(Insulin *
Glucose)/22,5[17].
Figure1.CD68‐stainingofadiposetissuemacrophages(ATMs).
A) Solitary ATM, B) Crown‐like structure (CLS): a well‐organized collection of macrophages
surroundinganadipocyte,whichisamarkerofadipocytedeath[34].CD68isalysosomalglycoprotein
that is implicated in endocytosis and lysosomal migration of macrophages. Note the difference in
macrophagesizeandgranulomatousaspectbetweenasolitaryATMandATMsformingaCLS.
Statisticalanalyses
Differences in descriptive characteristics between COPD patients and healthy
controlsweretestedusingindependent‐samplest‐testforcontinuousvariables
and χ2 for categorical variables. In case of a significantly skewed distribution
according to the Shapiro‐Wilk test, natural logarithmic transformation was
applied and for these variables the geometrical mean and 95% confidence
intervals (CI) are presented. Analyses of adipose tissue mRNA expression and
systemicadipocytokineswerecorrectedformultiplecomparisonsbyusingfalse
discovery rate method [18]. Correlations were tested using Pearson’s r and
Spearman’s ρ. Analyses were performed in PASW Statistics 17.0 (SPSS inc.,
Chicago,IL).DifferencesweredeclaredtobestatisticallysignificantatP<.05.
Results
COPDpatienthadmild‐to‐moderateairflowobstruction(Table1).FFMIandFMI
were not different between COPD patientsand healthy controls.COPD patients
had slightly higher fasting plasma glucose, insulin and HOMA‐IR scores
68
CHAPTER4
Table1.Maincharacteristicsofthestudypopulation.
Healthy TotalCOPD Desaturating
Non‐
controls
group
COPDpatients desaturating
(n=15)
(n=28)
(n=12) COPDpatients
(n=16)
Demographics
Sex,m/f
9/6
17/11
7/5
10/6
Age,y
65(6)
65(7)
65(6)
66(7)
Smokingstatus
Current,n(%)
1(6.7)
10(36)
2(20)
8(50)
Former,n(%)
7(46.7)
18(64)a
10(80)
8(50)
Never,n(%)
7(46.7)
0(0)
0(0)
0(0)
Pulmonaryfunction
FEV1,%pred
113(15)
58(16)a
57(19)
59(15)
FVC,%pred
120(17)
105(22)c
114(21)
98(17)
a
f
FEV1/FVC,%
74(5)
45(12)
39(9)
49(12)
DLCO,%pred
95(19)
51(16)a
48(16)
54(17)
a
SaO2 atrest,%
99(1)
96(2)
96(2)
97(1)
a
d
SaO2atVO2,max,%
98(1)
91(4)
87(3)
94(2)
PaO2atrest,kPa
12.1(1.1)
9.4(1.0)a
9.2(1.2)
9.6(0.8)
PaO2atVO2,max,kPag 13.1(1.3)
9.1(1.4)a
7.9(1.2)d
9.8(1.0)
Bodycomposition
2
e
BMI,kg/m 24.9(3.3)
24.9(2.9)
23.2(2.2)
26.1(2.7)
2
FFMI,kg/m
18.0(1.9)
17.6(1.7)
17.1(1.4)
17.9(1.9)
FMI,kg/m2
7.0(3.3)
7.2(2.6)
6.0(2.1)f
8.1(2.7)
Totalbodyfat,%
28.1(10.3) 29.2(8.2)
26.5(7.6)
31.3(8.1)
h,i
Cardiometabolicprofile
Glucose,mmol/L
5.3(5.1‐5.6) 5.9(5.6‐6.1)b
5.7 (5.3‐6.1) 6.0(5.7‐6.3)
Insulin,pmol/L
29(23‐37) 48(36‐64)c
30 (23‐37)d
70(46‐106)
HOMA‐IR
0.8 (0.5‐1.1) 1.5(1.0‐2.3)c 0.8(0.5‐1.3)d 2.6(1.6‐4.2)
NT‐proBNP,pmol/L 6.6(4.4‐9.8) 9.6(6.8‐13.5) 15.2(10.4‐22.2)f 6.6(4.1‐10.8)
Systemicinflammationh,i
CRP,mg/L
0.95(0.57‐1.59)2.05(1.31‐3.21)c 1.62(0.95‐2.76) 2.48(1.19‐5.15)
forced expiratory volume in 1s; FFMI, fat‐free mass index; FMI, fat mass index; FVC, forced vital
capacity; HOMA‐IR, homeostatic model assessment insulin resistance. Data are mean (SD) unless
indicated otherwise. aP<.001, bP<.01, cP<.05 compared to healthy controls. dP≤.001, eP<.01, fP<.05
comparedtoCOPDNDpatients. gAvailablefor7COPDD,13COPDNDand11healthycontrols. hNoplasma
availableforn=1COPDpatient.iGeometricmean(95%confidenceinterval).
69
than healthy controls (P≤.021) and CRP was also modestly but significantly
higherinCOPDpatients(P=.031)reflectingalow‐gradesystemicinflammatory
state.CRPandHOMA‐IRwerepositivelycorrelated(Figure2).
Figure2.CorrelationbetweenCRPandHOMA‐IR.
Closed circles represent COPD patients, open circles represent
healthycontrols(Spearman’sρ=0.36,P=.018,n=44).
ComparisonsbetweenhealthycontrolsandthetotalCOPDgroup
Adipocytesizeandthecorrelationbetweenfatmassandadipocytesizewerenot
different between the two groups (Figure 3). None of the measured gene
expression levels were significantly different between COPD patients and
healthy controls after adjustment for multiple testing (Table 2). The solitary
ATM density in COPD patients was not different from healthy controls (30.4
[23.9‐38.6] vs. 23.7 [18.0‐31.3] ATMs/mm2, P=.19). CLS were found in n=12
COPDpatients(43%)andn=5healthycontrols(33%)(P=.75).Inthissubset,the
CLSdensitywasnotdifferentbetweenCOPDpatientsandhealthycontrols(0.15
[0.08‐0.28] vs. 0.09 [0.04‐0.19] CLS/mm2, P=.27). CLS density was positively
correlatedwithCD68geneexpression(r=0.66,P=.003),whileATMdensitywas
not(SupplementalFigure1).
Noneofthemeasuredsystemicadipokinelevelsweresignificantlydifferent
betweenCOPDpatientsandhealthycontrols(Table3).SystemicIL‐6levelswere
near detection limit in bothCOPD patients and healthy controls and no
differenceswerefoundbetweenthesetwogroups.MeasuresabovetheLLODof
systemicIL‐1β,IL‐8,IL‐10,IFN‐γandTNF‐αwerefoundinonlyn=1,n=24,n=4,
n=15 and n=2 subjects, respectively, without clear differences between COPD
70
CHAPTER4
Figure3.AdipocytesizeinCOPDpatientsandhealthycontrols.
A)ComparisonofadipocytesizebetweenCOPDpatients(n=28,blackbars)andhealthycontrols(n=15,
gray bars). Data are means±SE. No differences were observed (all P>.05). B) Similar correlations
between fat mass and median adipocyte size in COPD patients (closed circles, solid line, Pearson’s
r=0.41,P=.031,n=28)andhealthycontrols(opencircles,dashedline,Pearson’sr=0.63,P=.012,n=15).C)
Haematoxylin‐eosinstainedadiposetissuesectionsat200×magnificationillustratingthesimilarityin
adipocytesizebetweenCOPDpatientsandhealthycontrols.
patientsandhealthycontrols.Becauseoftheselownumbers,wedidnotinclude
theseinouranalyses.Fatmass,circulatingleptinandadiponectin,andadipose
tissueleptinandadiponectingeneexpressioncorrelatedinasimilarfashionin
71
Table 2. Comparison of adipose tissue mRNA levels (arbitrary units) between
COPDpatientsandhealthycontrols.
Healthy
TotalCOPD
controls(n=15) group(n=28)
Desaturating
Non‐
(n=12)
COPDpatients
(n=16)
GLUT1
0.17(0.12‐0.23) 0.17(0.14‐0.20)
0.17(0.13‐0.23) 0.17(0.13‐0.22)
HIF1α
0.18(0.16‐0.20) 0.19(0.18‐0.20)
0.19(0.17‐0.20) 0.19(0.17‐0.21)
VEGF‐A
0.16(0.11‐0.22) 0.15(0.13‐0.19)
0.18(0.12‐0.26) 0.14(0.11‐0.18)
CD34
0.14(0.07‐0.27) 0.16(0.13‐0.20)
0.17(0.13‐0.22) 0.15(0.11‐0.22)
PAI‐1
0.14(0.11‐0.18) 0.18(0.14‐0.24)
0.14(0.09‐0.22) 0.22(0.16‐0.30)
Adrenomedullin
0.16(0.10‐0.26) 0.23(0.19‐0.27)
0.22(0.17‐0.29) 0.23(0.18‐0.29)
Visfatin
0.19(0.09‐0.37) 0.22(0.17‐0.28)
0.21(0.14‐0.30) 0.23(0.16‐0.32)
Osteoprotegerin
0.16(0.09‐0.29) 0.23(0.19‐0.28)
0.19(0.14‐0.27) 0.26(0.20‐0.33)
IL‐8
0.14(0.08‐0.26) 0.15(0.11‐0.22)
0.11(0.06‐0.21) 0.19(0.12‐0.31)
MCP‐1
0.11(0.08‐0.14) 0.16(0.13‐0.20)
0.16(0.12‐0.22) 0.16(0.11‐0.22)
CD68
0.15(0.11‐0.19) 0.18(0.15‐0.22)
0.16(0.12‐0.22) 0.20(0.16‐0.25)
CD163
0.11(0.05‐0.21) 0.17(0.12‐0.23)
0.14(0.08‐0.26) 0.20(0.14‐0.28)
CD206
0.13(0.09‐0.21) 0.21(0.16‐0.27)
0.17(0.10‐0.28) 0.24(0.19‐0.31)
CD11b
0.16(0.10‐0.26) 0.18(0.12‐0.27)
0.20(0.13‐0.30) 0.16(0.08‐0.34)
Leptina
0.16(0.11)
0.17(0.08)
0.16(0.10)
0.19(0.05)
Adiponectina
0.19(0.08)
0.18(0.07)
0.19(0.09)
0.17(0.05)
PPAR‐γa
0.19(0.09)
0.16(0.06)
0.17(0.07)
0.15(0.05)
Chemerina
0.20(0.10)
0.19(0.08)
0.20(0.08)
0.18(0.08)
IL‐10a
0.30(0.11)
0.28(0.08)
0.29(0.07)
0.28(0.08)
Abbreviations:GLUT1,glucosetransporter1;HIF1α,hypoxiainduciblefactor1α;IL,interleukin;MCP‐
1, monocyte chemotactic protein‐1; PAI‐1, plasminogen activator inhibitor‐1; PPAR‐γ, peroxisome
proliferator‐activated receptor‐γ; VEGF‐A, vascular endothelial growth factor‐A. Data are geometric
meansand95%confidenceintervals,exceptindicatedotherwise.Independent‐samplest‐testswereused
toanalyzethedata,andPvalueswereinterpretedapplyingfalsediscoveryratecorrectionformultiple
testing.Noneofthecomparisonswerestatisticallysignificant.aMean(SD).
both COPD patients and healthy controls (Figure 4). Positive correlations
betweenfatmassandATMmarkerswerealsohighlycomparablebetweenCOPD
patientsandhealthycontrols(Figure5).NeitherinCOPDpatientsnorinhealthy
controls any significant correlations were found between fat mass and the
remainingsystemicadipokines(allP>.05,datanotshown).
72
CHAPTER4
Comparisonsbetweendesaturatingandnon‐desaturatingCOPDpatients
BMI (P=.007) and FMI (P=.040) were significantly lower in desaturating COPD
patients (COPDD). Non‐desaturating COPD patients (COPDND) had higher
FEV1/FVC (P=.028), higher HOMA‐IR (P≤.001) and lower NT‐proBNP (P=.010).
NomarkeddifferencesinadipocytesizewerenotedbetweenCOPDDandCOPDND
patients (data not shown). We found nosignificantdifferences between COPDD
and COPDND patients for all of the measured SAT gene expressions (Table 2),
systemicadipokinelevels(Table3),ATMdensity(30.6[21.3‐43.8]vs.30.2[21.2‐
43.2] ATMs/mm2, P=.96) and CLS density (0.10 [0.05‐0.20] vs. 0.20 [0.07‐0.56]
CLS/mm2,P=.22).
Based onthemedian CRP level in COPD patients, weperformedapost‐hoc
comparisonbetweenpatientswithhighCRP(n=14,CRP4.38[2.33‐8.24]mg/L)
and low CRP (n=13, CRP 0.90 [0.75‐1.08] mg/L) (Supplemental Tables 3‐5)
showing that high CRP patients had higher ATM density (P=.021, Figure 6).
Linear regression analysis further revealed that ATM density (ln transformed)
was significantly predicted by CRP status (low CRP=1, high CRP=2, β=0.61,
SE=0.23, P=.013) independent of BMI (β=‐0.04, SE=0.04, P=.29). In a separate
model CRP status (β=0.63, SE=0.22, P=.008) also predicted ATM density
independentoffatmass(β=‐0.03,SE=0.02,P=.07).
Discussion
We investigated whether ATI is enhanced in stable mild‐to‐moderate COPD
independent of age, sex, BMI and body composition. The second aim was to
investigate whether ATI was more pronounced in COPD patients with a
desaturation phenotype. Knowledge of potential adipose tissue dysfunction in
COPD may increase our understanding of the origin of enhanced systemic
inflammation in COPD and contribute to tailoring interventions. COPD patients
werecharacterizedbyaslightlyhigherCRPandhigherHOMA‐IRthatcouldbe
interrelated (as supported by a positive correlation). COPD patients tended to
havehigheradiposetissuegeneexpressionoftwomacrophagemarkers(MCP‐1
and CD206). We found that COPD patients with high CRP had greater ATM
infiltration and tended to have higher CD206 expression compared to low‐CRP
patients. Other markers of ATI were, however, not different between COPD
patients and healthy controls, and neither was the desaturation phenotype of
influenceonanyofthesemarkers.
As larger adipocytes have been shown to be more pro‐inflammatory than
smalleradipocytes[3],weexploredwhetherCOPDpatientswouldhaveenlarged
adipocytes, for which we did not find any evidence. In mild‐to‐moderate COPD
patients,Skybaetal.reportedapositiveassociationbetweenadipocytesizeand
BMI which was not associated with different gene expression of apoptosis
markers, suggesting that fat wasting is not associated with
73
Table3.ComparisonofsystemicadipokinelevelsbetweenCOPDpatientsaand
healthycontrols.
IL‐6,pg/ml
Leptin,ng/ml
Adiponectin,μg/ml
PAI‐1,μg/ml
Chemerin,μg/ml
Adipsin,μg/ml
Resistin,μg/ml
SAA‐1,μg/ml
CXCL10,ng/ml
MIF,ng/ml
MIP‐1α,pg/mlb
MCP‐1,ng/mlc
OPG,ng/mlc
Healthy
TotalCOPD
controls(n=15) group(n=28)
3.73(2.52‐5.52)
3.92(2.32‐6.63)
25.6(19.9‐32.9)
1.60(1.25‐2.05)
0.33(0.28‐0.39)
1.25(1.04‐1.51)
3.16(2.54‐3.93)
8.04(6.88‐9.39)
1.55(1.22‐1.96)
2.51(1.92‐3.28)
37.5(20.3‐69.1)
0.95(0.48)
1.18(0.45)
5.04(3.12‐8.15)
5.19(3.73‐7.22)
19.7(15.5‐25.1)
1.72(1.24‐2.39)
0.33(0.28‐0.39)
1.18(0.96‐1.46)
2.86(2.40‐3.41)
7.55(6.67‐8.54)
1.73(1.36‐2.20)
2.30(1.71‐3.09)
55.4(41.2‐74.4)
1.05(0.39)
1.10(0.27)
Non‐
Desaturating
COPDpatients
(n=12)
(n=16)
5.63(1.86‐17.1)
4.03(2.40‐6.77)
23.9(14.5‐39.3)
2.36(1.17‐4.73)
0.38(0.28‐0.53)
1.23(0.87‐1.74)
3.12(2.22‐4.39)
8.06(6.27‐10.4)
2.06(1.32‐3.23)
2.35(1.87‐2.95)
45.0(23.7‐85.3)
1.03(0.41)
1.05(0.25)
4.62(3.38‐6.30)
6.36(4.05‐9.99)
16.9(13.6‐20.9)
1.34(1.05‐1.71)
0.29(0.25‐0.34)
1.14(0.84‐1.54)
2.68(2.19‐3.28)
7.17(6.31‐8.14)
1.51(1.15‐1.98)
2.26(1.33‐3.86)
64.0(47.0‐87.0)
1.06(0.38)
1.15(0.28)
Abbreviations: CXCL10, C‐X‐C motif chemokine 10; IL‐1RA, interleukin‐1 receptor antagonist; IL‐6,
interleukin‐6; MCP‐1, monocyte chemotactic protein‐1; MIF, macrophage migration inhibiting factor;
MIP‐1α, macrophage inflammatory protein‐1α; OPG, osteoprotegerin; PAI‐1, plasminogen activator
inhibitor‐1;SAA‐1,serumamyloidA1.Dataaregeometricmeansand95%confidenceintervals,except
indicatedotherwise. aNoplasmaavailableforn=1COPDpatient. bValidmeasuresabovethelowerlimit
of detection available for n=22 COPD patients and n=13 healthy controls. cMean (SD). None of the
comparisonswerestatisticallysignificant.
adipocyte death but rather with adipocyte atrophy [11]. In line, we report a
positive relation between fat mass and adipocyte size in COPD, which was
similarinhealthycontrols.Thissuggeststhatthebehaviorofadipocyteatrophy
andhypertrophywithchangingfatmassfollowsaphysiologicalpatterninCOPD
patients.
ATMsarebelievedtobemediatorsofATI,soaquantificationofATMswould
provideanindicationofATI.WhereaswefoundnodifferencesinATMinfiltration
between COPD patients and healthy controls, ATM infiltration was higher in
high‐CRP COPD patients compared to low‐CRP patients, suggesting a relation
betweenlow‐gradesystemicinflammationandATI.Interestingly,wefoundthat
CD68mRNAlevelswerenotcorrelatedwiththedensityofsolitaryATMsbuta
correlation was found with the density of CLS. As CD68 is a lysosomal
glycoprotein that is implicated in endocytosis and lysosomal migration of
macrophages, our findings suggest that ATMs involved in a CLS express
considerablymoreCD68comparedtoresiding/restingsolitaryATMs(Figure1).
Inlinewithpreviouspublications[11,19],wefoundthatfatmasswaspositively
74
CHAPTER4
Figure 4. Correlations between fat mass, adipose tissue mRNA and systemic
concentrations of leptin and adiponectin in COPD patients (closed circles) and
healthycontrols(opencircles).
A)Correlationbetweenfatmassandsystemicleptinconcentration,COPDpatients:Spearman’sρ=0.62,
P=.001,n=27,healthycontrols:Spearman’sρ=0.58,P=.029,n=14,B)Correlationbetweenadiposetissue
leptinmRNAexpressionandsystemicleptinconcentration,COPDpatients:Spearman’sρ=0.50,P=.009,
n=27,healthycontrols:Spearman’sρ=0.63,P=.016,n=14,C)Correlationbetweenfatmassandsystemic
adiponectin concentration, COPD patients: Spearman’s ρ=‐0.05, P=.80, n=27, healthy controls:
Spearman’sρ=‐0.14,P=.63,n=15,D)CorrelationbetweenadiposetissueadiponectinmRNAexpression
and systemic adiponectin concentration, COPD patients: Spearman’s ρ=0.41, P=.033, n=27, healthy
controls:Spearman’sρ=0.68,P=.005,n=15.
correlatedwithmarkersofATMs,whichwashighlycomparablebetweenCOPD
patientsandhealthycontrols.Thishasbeenwell‐acceptedtobeacharacteristic
ofobesityandhasbeenlinkedtothedevelopmentofIRviaATI[19].
Interestingly, despite the absence of greater ATM density, COPD patients
were characterized by a higher HOMA‐IR compared to the healthy controls,
suggesting that IR in COPD patients is not necessarily associated with greater
ATMinfiltration.Likewise,COPDNDpatientshadconsiderablyhigherinsulinand
HOMA‐IRcomparedtoCOPDDpatients,butnodifferencesinATMinfiltrationwere
found. It must, however, be mentioned that we did not perform a
hyperinsulinemiceuglycemicclamptoverifyIR.
75
Figure 5. Correlations between fat mass and markers of adipose tissue
macrophages in COPD patients (closed circles) and healthy controls (open
circles).
A) Correlation between fat mass and adipose tissue CD163 mRNA expression, COPD patients:
Spearman’s ρ=0.36, P=.060, n=28, healthy controls: Spearman’s ρ=0.66, P=.007, n=15, B) Correlation
betweenfatmassandadiposetissueCD68mRNAexpression,COPDpatients:Spearman’sρ=0.42,P=.025,
n=28,healthycontrols:Spearman’sρ=0.66,P=.007,n=15,C)Correlationbetweenfatmassandadipose
tissue CD206 mRNA expression, COPD patients: Spearman’s ρ=0.60, P=.001, n=28, healthy controls:
Spearman’s ρ=0.64, P=.010, n=15, D) Correlation between fat mass and adipose tissue MCP‐1 mRNA
expression, COPD patients: Spearman’s ρ=0.41, P=.033, n=28, healthy controls: Spearman’s ρ=0.30,
P=.28,n=15.
Animal and in vitro studies indicate that hypoxia may lead to ATI. For
example, hypoxia increased the activity of the pro‐inflammatory transcription
factor Nuclear Factor‐ĸB and the TNF‐α gene promoter in murine adipocytes
[20],anditderegulatestheproductionofadiponectinandPAI‐1[21].Inobesity,
ithasbeenpostulatedthatadipocytehypertrophycombinedwithimpairedneo‐
vascularisation may ultimately lead to an unbridgeable diffusion distance for
oxygenbothonthetissuelevelaswellasonthecellularlevel[5].Indeed,obese
mice had hypoxic patches in the adipose tissue which was associated with ATI
[20, 22]. However, human data on adipose tissue hypoxia as underlying ATI is
scarce.Recently,Goossensetal.challengedtheconceptofadiposetissuehypoxia
in humans by showing that obese insulin resistant men were characterized by
76
CHAPTER4
adiposetissuehyperoxiaratherthanhypoxia[23].Wefoundthatinthepresence
ofamildhypoxaemiaintheCOPDpatientsrelativetothehealthycontrols(i.e.,a
meandifferenceinPaO2of2.7kPa),nocleardifferencescouldbeobservedwith
regard to markers of adipose tissue hypoxia or ATI. Assuming that COPDD
patientswouldalsodesaturateduringdailylivingactivities[6],weshowthatthe
desaturating COPD phenotype is not associated with either of these outcomes.
Additionally, adipocyte size and (neo‐) vascularisation markers were
comparableacrossthecurrentstudygroups.Collectively,ourdatasuggestthat
neithermildhypoxaemianoradesaturationphenotypeisassociatedwithATIin
COPDpatients.
Figure6.AdiposetissuemacrophagesinCOPD
patientswithlowversushighCRP.
Datapointsaregeometricalmeans,errorbarsrepresentthe
95% confidence interval. Data were tested using
independent‐samplest‐test.
Several authors have claimed disturbances in systemic adipokine levels in
COPD patients [24‐29]. Takabatake et al. [28] showed lower systemic leptin
levels in underweight COPD patients compared to normal‐weight non‐COPD
controlsandEkeretal.[30]foundlowersystemicleptinlevelsinCOPDpatients
withdecreasedFFMcomparedtoBMI‐andsex‐matchedhealthycontrols.Inour
study, none of the measured systemic adipokines (including leptin and
77
adiponectin) revealed different levels between COPD patients and matched
healthycontrols,indicatingtheimportanceofadjustingforfatmass.Indeed,we
found a strong positive correlation between fat mass and circulating leptin
levels,whichwasfoundtobesimilarinCOPDpatientsandhealthycontrols.In
addition, adipose tissue expression of leptin and adiponectin were positively
correlated with systemic leptin and adiponectin levels, respectively, also in a
similarfashioninCOPDpatientsasinhealthycontrols.
Recently, important distinct functions and characteristics of SAT versus
visceral adipose tissue (VAT) have been reported. For example, the relative
contribution of SAT compared to VAT to systemic adipokine levels may be
limited[31]anddifferencesexistinneovascularisationwithexpandingfatmass
whichmayexertdifferentmetaboliceffectsinSATandVAT[32].Also,expanding
VAT is generally associated with increased cardiovascular risk whereas
expanding SAT may even lower cardiovascular risk especially in women [33].
Therefore, it would be interesting to explore a potential VAT dysfunction in
COPD patients in future studies and compare it to SAT function. In addition,
studyingpotentialadiposetissuedysfunctioninmoresevereCOPDpatientsand
during acute exacerbations may add to optimizing nutritional and lifestyle
interventions.
In conclusion, markers of ATI and systemic adipokine levels in clinically
stable mild‐to‐moderate COPD patients who are characterized by low‐grade
systemic inflammation, higher HOMA‐IR and mild relative hypoxaemia, were
comparabletoage‐,sex‐,BMI‐andbodycomposition‐matchedhealthycontrols.
OurstudyprovidesafirstindicationforapossibleroleofATMsinthesystemic
inflammatoryresponseinCOPDthatwarrantsfurtherinvestigation.
78
CHAPTER4
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
MancusoP.Obesityandlunginflammation.JApplPhysiol.2010;108(3):722‐8.
SkurkT,Alberti‐HuberC,HerderC,HaunerH.Relationshipbetweenadipocytesizeandadipokineexpression
andsecretion.JClinEndocrinolMetab.2007;92(3):1023‐33.
SchiffelersSL,BlaakEE,BaarendsEM,VanBaakMA,SarisWH,WoutersEF,ScholsAM.beta‐Adrenoceptor‐
mediatedthermogenesis and lipolysisin patientswith chronic obstructive pulmonary disease.Am J Physiol
EndocrinolMetab.2001;280(2):E357‐64.
Trayhurn P, WangB, Wood IS. Hypoxia in adipose tissue: a basis for the dysregulation of tissue function in
obesity?BrJNutr.2008;100(2):227‐35.
Soguel Schenkel N, Burdet L, de Muralt B, Fitting JW. Oxygen saturation during daily activities in chronic
obstructivepulmonarydisease.EurRespirJ.1996;9(12):2584‐9.
BoltonCE,EvansM,IonescuAA,EdwardsSM,MorrisRH,DunseathG,LuzioSD,OwensDR,ShaleDJ.Insulin
resistanceandinflammation‐AfurthersystemiccomplicationofCOPD.Copd.2007;4(2):121‐6.
Torres‐LealFL,Fonseca‐AlanizMH,RogeroMM,TirapeguiJ.Theroleofinflamedadiposetissueintheinsulin
resistance.CellBiochemFunct.2011;28(8):623‐31.
MorrisDL,SingerK,LumengCN.Adiposetissuemacrophages:phenotypicplasticityanddiversityinleanand
obesestates.CurrOpinClinNutrMetabCare.2011;14(4):341‐6.
TkacovaR,UkropecJ,SkybaP,UkropcovaB,PobehaP,KurdiovaT,JoppaP,KlimesI,TkacI,GasperikovaD.
Increased adipose tissue expression of proinflammatory CD40, MKK4 and JNK in patients with very severe
chronicobstructivepulmonarydisease.Respiration.2011;81(5):386‐93.
SkybaP,UkropecJ,PobehaP,UkropcovaB,JoppaP,KurdiovaT,StroffekovaK,BrusikM,KlimesI,TkacI,etal.
Metabolic phenotype and adipose tissue inflammation in patients with chronic obstructive pulmonary
disease.MediatorsInflamm.2010;2010:173498.
FuldaS,LinseisenJ,WolframG,HimmerichS,GedrichK,PollmacherT,HimmerichH.Leptinplasmalevelsin
the general population: influence of age, gender, body weight and medical history. Protein Pept Lett.
2010;17(11):1436‐40.
Boyne MS, Bennett NR, Cooper RS, Royal‐Thomas TY, Bennett FI, Luke A, Wilks RJ, Forrester TE. Sex‐
differences in adiponectin levels and body fat distribution: longitudinal observations in Afro‐Jamaicans.
DiabetesResClinPract.2010;90(2):e33‐6.
ATS/ACCPStatementoncardiopulmonaryexercisetesting.AmJRespirCritCareMed.2003;167(2):211‐77.
VandesompeleJ,DePreterK,PattynF,PoppeB,VanRoyN,DePaepeA,SpelemanF.Accuratenormalization
2002;3(7):RESEARCH0034.
SchipperHS,deJagerW,vanDijkME,MeerdingJ,ZelissenPM,AdanRA,PrakkenBJ,KalkhovenE.Amultiplex
immunoassayforhumanadipokineprofiling.ClinChem.2010;56(8):1320‐8.
WallaceTM,LevyJC,MatthewsDR.UseandabuseofHOMAmodeling.DiabetesCare.2004;27(6):1487‐95.
BenjaminiY,HochbergY.Controllingthefalsediscoveryrate:apracticalandpowerfulapproachtomultiple
testing.JournaloftheRoyalStatisticalSociety.1995;57(1):289‐300.
Red Eagle A, Chawla A. In obesity and weight loss, all roads lead to the mighty macrophage. J Clin Invest.
2010;120(10):3437‐40.
YeJ,GaoZ,YinJ,HeQ.Hypoxiaisapotentialriskfactorforchronicinflammationandadiponectinreductionin
adiposetissueofob/obanddietaryobesemice.AmJPhysiolEndocrinolMetab.2007;293(4):E1118‐28.
Chen B, Lam KS, Wang Y, Wu D, Lam MC, Shen J, Wong L, Hoo RL, Zhang J, Xu A. Hypoxia dysregulates the
production of adiponectin and plasminogen activator inhibitor‐1 independent of reactive oxygen species in
adipocytes.BiochemBiophysResCommun.2006;341(2):549‐56.
HosogaiN,FukuharaA,OshimaK,MiyataY,TanakaS,SegawaK,FurukawaS,TochinoY,KomuroR,Matsuda
M, et al. Adipose tissue hypoxia in obesity and its impact on adipocytokine dysregulation. Diabetes.
2007;56(4):901‐11.
Goossens GH, Bizzarri A, Venteclef N, Essers Y, Cleutjens JP, Konings E, Jocken JW, Cajlakovic M, RibitschV,
ClementK,etal.Increasedadiposetissueoxygentensioninobesecomparedwithleanmenisaccompaniedby
insulinresistance,impairedadiposetissuecapillarization,andinflammation.Circulation.2011;124(1):67‐76.
BreyerMK,RuttenEP,VernooyJH,SpruitMA,DentenerMA,vanderKallenC,VangreevenbroekMM,Wouters
EF.Genderdifferencesintheadiposesecretomesysteminchronicobstructivepulmonarydisease(COPD):A
pivotalroleofleptin.RespirMed.2011;105(7):1046‐53.
Wang QY, Zhang H, Yan X, Kang J, Yu RJ. [Serum resistin and leptin in patients with chronic obstructive
pulmonary disease and their relationship to nutritional state]. Zhonghua Jie He He Hu Xi Za Zhi.
2005;28(7):445‐7.
Poulain M, Doucet M, Drapeau V, Fournier G, Tremblay A, Poirier P, Maltais F. Metabolic and inflammatory
profileinobesepatientswithchronicobstructivepulmonarydisease.ChronRespirDis.2008;5(1):35‐41.
Tomoda K, Yoshikawa M, Itoh T, Tamaki S, Fukuoka A, Komeda K, Kimura H. Elevated circulating plasma
adiponectininunderweightpatientswithCOPD.Chest.2007;132(1):135‐40.
TakabatakeN,NakamuraH,AbeS,HinoT,SaitoH,YukiH,KatoS,TomoikeH.Circulatingleptininpatients
withchronicobstructivepulmonarydisease.AmJRespirCritCareMed.1999;159(4Pt1):1215‐9.
Yuan Y, Wang Z, Liu C. [Preliminary investigation of effect of serum leptin on nutritional state of COPD
patients].ZhonghuaJieHeHeHuXiZaZhi.2000;23(5):292‐5.
79
30. EkerS,AyazL,TamerL,UlubasB.Leptin,visfatin,insulinresistance,andbodycompositionchangeinchronic
obstructivepulmonarydisease.ScandJClinLabInvest.2010;70(1):40‐4.
31. HockingSL,WuLE,GuilhausM,ChisholmDJ,JamesDE.Intrinsicdepot‐specificdifferencesinthesecretomeof
adipose tissue, preadipocytes, and adipose tissue‐derived microvascular endothelial cells. Diabetes.
2010;59(12):3008‐16.
32. GealekmanO,GusevaN,HartiganC,ApothekerS,GorgoglioneM,GuravK,TranKV,StraubhaarJ,NicoloroS,
Czech MP, et al. Depot‐specific differences and insufficient subcutaneous adipose tissue angiogenesis in
humanobesity.Circulation.2011;123(2):186‐94.
33. ZafonC.Fatandaging:ataleoftwotissues.CurrAgingSci.2009;2(2):83‐94.
34. Cinti S, Mitchell G, Barbatelli G, Murano I, Ceresi E, Faloia E, Wang S, Fortier M, Greenberg AS, Obin MS.
Adipocytedeathdefinesmacrophagelocalizationandfunctioninadiposetissueofobesemiceandhumans.J
LipidRes.2005;46(11):2347‐55.
35. QuanjerPH,TammelingGJ,CotesJE,PedersenOF,PeslinR,YernaultJC.Lungvolumesandforcedventilatory
flows.ReportWorkingPartyStandardizationofLungFunctionTests,EuropeanCommunityforSteelandCoal.
OfficialStatementoftheEuropeanRespiratorySociety.EurRespirJSuppl.1993;16:5‐40.
80
CHAPTER4
Supplementalmaterial
Detailedmethodology
Clinically stable COPD patients were recruited from the outpatient clinic of the
Maastricht University Medical Centre (MUMC+, Maastricht, The Netherlands)
andusingadvertisementsinlocalnewspapers.Patientswereexcludedincaseof
long‐term oxygen therapy, acute exacerbation with hospital admission in the
past eight weeks, oral steroid use, rehabilitation in the past six months,
scheduled rehabilitation and α1‐antitrypsin deficiency. COPD patients with
known chronic heart failure, cancer treatment in the past three years,
sarcoidosis,pulmonaryembolism,obstructiveorcentralsleepapneasyndrome,
diabetes mellitus, hepatic or renal failure, rheumatoid arthritis, M. Bechterew
and inflammatory bowel disease were excluded from this study. Subjects
performed an incremental exercise test (see below for methodology) to
determine exercise‐induced desaturation defined as a fall in oxygen saturation
(SaO2) of ≥4% [14]. This study included 28 COPD patients of whom 12
desaturated(COPDD,mean±SDΔSaO2‐8.5±2.2%)and16didnot(COPDND,ΔSaO2
‐2.1±1.7%). Fifteen healthy controls were recruited via advertising in local
newspapers. The absence of co‐morbidities in the healthy controls (including
asthma and COPD) was ascertained based on a physician’s diagnosis, lung
function,self‐reportedhistoryand/ortreatment.Writteninformedconsentwas
obtainedfromallsubjectsandtheethicalreviewboardoftheMUMC+approved
thestudy(MEC08‐2‐059).
Pulmonaryfunctionandincrementalexercisetest
Pulmonary function testing included forced spirometry and single breath
diffusion capacity measurement (Masterlab, Jaeger, Würzburg, Germany).
Instrumentswerecalibratedtwiceaday.Allvaluesobtainedwereexpressedasa
percentageofreferencevalues[35].Allsubjectsperformedanincrementalcycle
ergometrytestaccordingtointernationalstandards[14].Afterthreeminutesof
unloadedcycling,workloadwasincreasedby10Watteveryminuteinpatients.
Incontrolsubjects,theloadwasincreasedwith15‐25Watt·min‐1,dependingon
relativefitnessofthesubject,soastoleadtofatiguein10‐12minutes.Subjects
were encouraged to cycle at 60 rpm until exhaustion. SaO2 during the test was
monitored by pulse oximetery. Arterial blood gas analysis (Blood gas analyzer
865, Chiron Diagnostics, Emeryville, CA, USA) was performed at rest and at
VO2maxfromaradialarterypuncture.
81
Cardiometabolicandsystemicinflammatoryprofile
Afteranovernightfast,venousbloodwasdrawninthemorningandplasmaand
serum were stored at ‐80°C until further analysis. Plasma concentrations of 19
adipo(cyto)kines were analyzed with a multiplex immunoassay as described in
detail elsewhere [16]. The lower limits of detection (LLOD) are presented in
Table E3. For IL‐6, 28% (n=7 COPD patients and n=5 healthy controls) of the
data set was below the last linear point of the standard curve (3.6 pg/ml) but
above the LLOD and therefore extrapolated data was used. Serum C‐reactive
protein (CRP) was determined with the CardioPhase® high‐sensitive CRP kit
(Siemens Healthcare Diagnostics Inc., Newark, NJ) (LLOD 0.18 mg/L). Glucose
was measured on a SensoStar GL 30 (DiaSys Diagnostics Systems GmbH,
Holzheim, Germany) and insulin was measured by the AutoDELFIA® time‐
resolvedfluoroimmunoassay(PerkinElmer,Turku,Finland).Homeostaticmodel
assessment (HOMA) wasapplied to estimate IR by HOMA‐IR=(Fasting insulin *
Fastingglucose)/22,5[17].SerumN‐terminalpro‐BrainNatriureticPeptide(NT‐
proBNP) was measured using an electrochemiluminescence immunoassay
(RocheDiagnostics,Woerden,TheNetherlands).
SupplementalTable1.PulmonarymedicationinCOPDpatients(n=28).
82
Pulmonarymedication
n
%
Long‐actingβ2agonist(LABA)
Short‐actingβ2agonist
Long‐actinganticholinergic
Short‐actinganticholinergic
Inhaledcorticosteroids(ICS)
LABA+ICS
Anyβ2agonist
4
10
18
4
4
13
20
14
36
64
14
14
46
71
CHAPTER4
SupplementalTable2.PrimersequencesusedforRT‐qPCR.
Forward5’‐3’
Reverse5’‐3’
Targetgenes
Leptin
CGGAGAGTACAGTGAGCCAAGA
CGGAATCTCGCTCTGTCATCA
Adiponectin
CCCAAAGAGGAGAGAGGAAGCT
GCCAGAGCAATGAGATGCAA
Chemerin
AAGTCAGGCCCAATGGGAGG
CCAACCGGCCCAGAACTTTGT
PAI‐1
GAAAGTGAAGATCGAGGTGAACGAGA
CATGCGGGCTGAGACTATGACA
Adrenomedullin ACTTGGCAGATCACTCTCTTAGCA
TCAGGGCGACGGAAACC
CD163
GCAAGAATTCATCTCCCGGTATTG
AGACAGAGACAGCGGCTTGCA
CD206
TCTGGAATGTCTTCGAATGGGTTC
GGCTCAACCCGATATGACAGAAAA
CD11b
ACACGGGATCGGCTAAGAGAAGGA
CGGGAATGTGGGCGGC
Visfatin
GGAAACCCTCTTGACACTGTGTTAAAGG
AAGATAAGGTGGCAGCAACTTGTAACC
Osteoprotegerin TTGGAGTGGTGCAAGCTGGAA
ACCCATCTGGACATCTTTTGCAAA
PPAR‐γ
CGGGCCCTGGCAAAAC
AAGATCGCCCTCGCCTTT
IL‐8
CACTGCGCCAACACAGAAAT
GAGCTCTCTTCCATCAGAAAGC
IL‐10
ATGGTGGCGCGCACCTG
CCTGGGTTCAAGCAATTCTCTTGC
VEGF‐A
CCAGGCCCTCGTCATTG
AAGGAGGAGGGCAGAATCAT
CD34
TAACGCCCTCCGCCTTTGG
GCGCCTCTCCTTGGCTGCTA
MCP1
AGCAGCAAGTGTCCCAAAGAAGCT
CCTTGGCCACAATGGTCTTGAA
CD68
GGACATTCTCGGCTCAGAAT
GCCTGGTAGGCGATGGGCG
GLUT‐1
TCTGGGCTGCCGGGTTCTAG
TTTGCAGGCTCCCACAGGC
HIF‐1α
TGAACATAAAGTCTGCAACATGGA
TGAGGTTGGTTACTGTTGGTATCATATA
Housekeepinggenes
GAPDH
GCACCACCAACTGCTTAGCA
TGGCAGTGATGGCATGGA
β2‐microglobulin TGACTTTGTCACAGCCCAAGATA
AATGCGGCATCTTCAAACCT
RPL13A
TTGAGGACCTCTGTGTATTTGTCAA
CCTGGAGGAGAAGAGGAAAGAGA
Abbreviations: GAPDH, glyceraldehydes 3‐phosphate dehydrogenase; GLUT‐1, glucose transporter‐1;
HIF‐1α, hypoxia inducible factor‐1α; MCP‐1, monocyte chemotactic protein‐1; PAI‐1, plasminogen
activatorinhibitor‐1;PPAR‐γ,peroxisomeproliferator‐activatedreceptor‐γ;RPL13A,Ribosomalprotein
13A;VEGF‐A,vascularendothelialgrowthfactor‐A.
83
Supplemental Table 3. Lower limits of detection (LLOD) of plasma
adipo(cyto)kines.
Adipo(cyto)kines
IL‐1RA
IL‐1β
IL‐6
IL‐10
TNF‐α
IFN‐γ
MIF
CCL2(MCP‐1)
CCL3(MIP1α)
CXCL8(IL‐8)
CXCL10(IP10)
Leptin
OPG
Chemerin
Adipsin(ComplFactorD)
Resistin
SAA‐I
PAI‐I
Adiponectin
LLOD(pg/ml)
0.7
0.5
0.8
0.8
0.4
1.4
1.4
0.4
3.1
1.8
0.3
0.3
0.7
31.7
0.4
12.8
832
88.2
36.4
Abbreviations: CCL, C‐C motif ligand; CXCL, C‐X‐C motif ligand; IL, interleukin; MCP‐1, monocyte
chemotactic protein‐1; MIF, macrophage migration inhibiting factor; MIP‐1α, macrophage
inflammatoryprotein‐1α;OPG,osteoprotegerin;PAI‐1,plasminogenactivatorinhibitor‐1;SAA‐1,serum
amyloidA1.
84
CHAPTER4
SupplementalTable4.ComparisonbetweenCOPDpatientswithlowCRP(i.e.
CRP<median)versushighCRP(i.e.CRP≥median).
Demographics
Sex,m/f
Age,y
Smokingstatus
Current,n(%)
Former,n(%)
Never,n(%)
Pulmonaryfunction
FEV1,%pred
FVC,%pred
FEV1/FVC,%
DLCO,%pred
SaO2atrest,%
SaO2atVO2,max,%
PaO2atrest,kPa
PaO2atVO2,max,kPaa
Bodycomposition
BMI,kg/m2
FFMI,kg/m2
FMI,kg/m2
Totalbodyfat,%
Cardiometabolicprofileb,c
Glucose,mmol/L
Insulin,pmol/L
HOMA‐IR
NT‐proBNP,pmol/L
Systemicinflammationb,c
CRP,mg/L
LowCRPCOPD
patients(n=13)
HighCRPCOPDpatients
(n=14)
8/5
66(7)
8/6
65(7)
6(46)
7(54)
0(0)
3(21)
11(79)
0(0)
57(18)
110(19)
42(13)
48(15)
97(2)
92(5)
9. 6(1.1)
9.4(1.9)
60(16)
100(24)
47(11)
54(18)
96(1)
91(4)
9.4 (1.0)
8.9 (1.1)
24.0(2.7)
17.1(1.6)
6.8(3.0)
28(10)
25.5(3.1)
17.9(1.9)
7.7(2.4)
31(7)
5.7(5.4‐6.0)
47(28‐78)
1.38(0.68‐2.78)
9.0(5.9‐13.8)
6.0(5.7‐6.4)
48(33‐71)
1.68(1.02‐2.77)
10.2(5.7‐18.2)
0.90(0.75‐1.08)
4.38(2.33‐8.24)d
forced expiratory volume in 1s; FFMI, fat‐free mass index; FMI, fat mass index; FVC, forced vital
capacity; HOMA‐IR, homeostatic model assessment insulin resistance. Data are mean (SD) unless
indicatedotherwise.Continuousvariablesweretestedusingindependent‐samplest‐testandcategorical
variables were tested using the χ2‐test. aAvailable for 20 patients. bNo plasma available for n=1 COPD
patient.cGeometricmean(95%confidenceinterval).dP<.001.
85
Supplemental Table 5. Comparison of adipose tissue mRNA levels (arbitrary
units)betweenlowandhighCRPCOPDpatients.
GLUT1
HIF1α
VEGF‐A
CD34
PAI‐1
Adrenomedullin
Visfatin
Osteoprotegerin
IL‐8
MCP‐1
CD68
CD163
CD206
CD11b
Leptina
Adiponectina
PPAR‐γa
Chemerina
IL‐10a
LowCRPCOPDpatients
(n=13)
HighCRPCOPDpatients
(n=14)
0.17(0.13‐0.23)
0.19(0.17‐0.20)
0.18(0.14‐0.24)
0.19(0.12‐0.28)
0.16(0.10‐0.25)
0.22(0.17‐0.28)
0.19(0.13‐0.28)
0.19(0.14‐0.26)
0.14(0.07‐0.26)
0.14(0.09‐0.20)
0.16(0.12‐0.20)
0.15(0.09‐0.25)
0.16(0.11‐0.24)
0.13(0.05‐0.30)
0.17(0.09)
0.20(0.06)
0.18(0.05)
0.21(0.07)
0.29(0.08)
0.16(0.12‐0.21)
0.19(0.17‐0.21)
0.14(0.11‐0.19)
0.14(0.11‐0.17)
0.20(0.14‐0.30)
0.24(0.18‐0.31)
0.23(0.16‐0.33)
0.26(0.20‐0.35)
0.16(0.10‐0.25)
0.18(0.14‐0.23)
0.20(0.16‐0.25)
0.19(0.12‐0.30)
0.26(0.19‐0.37)
0.23(0.16‐0.32)
0.19(0.06)
0.16(0.07)
0.15(0.06)
0.17(0.08)
0.28(0.07)
Abbreviations: GLUT1, glucose transporter 1; HIF1α, hypoxia inducible factor 1α; VEGF‐A, vascular
endothelial growth factor‐A; PAI‐1, plasminogen activator inhibitor‐1; MCP‐1, monocyte chemotactic
protein‐1, PPAR‐γ, peroxisome proliferator‐activated receptor‐γ. Data are geometric means and 95%
confidenceintervals,exceptindicatedotherwise.Independent‐samplest‐testswereusedtoanalyzethe
data,andPvalueswereinterpretedapplyingfalsediscoveryratecorrectionformultipletesting.aMean
(SD).
86
CHAPTER4
Supplemental Table 6. Comparison of systemic adipokine levels between low
andhighCRPCOPDpatients.
IL‐6,pg/ml
Leptin,ng/ml
Adiponectin,μg/ml
PAI‐1,μg/ml
Chemerin,μg/ml
Adipsin,μg/ml
Resistin,μg/ml
SAA‐1,μg/ml
CXCL10,ng/ml
IL‐1RA,ng/ml
MIF,ng/ml
MIP‐1α,pg/mla,b
MCP‐1,ng/mlb
OPG,ng/mlb
LowCRPCOPDpatients
(n=13)
4.34(3.03‐6.22)
4.98(2.89‐8.59)
22.2(17.6‐28.2)
1.90(0.95‐3.77)
0.35(0.26‐0.47)
1.17(0.79‐1.74)
2.59(1.95‐3.44)
7.73(6.27‐9.53)
1.89(1.40‐2.56)
1.66(1.19‐2.32)
2.60(1.70‐3.98)
73(50)
1.17(0.38)
1.10(0.27)
HighCRPCOPDpatients
(n=14)
5.79(2.29‐14.6)
5.39(3.40‐8.55)
17.6(11.3‐27.2)
1.57(1.22‐2.01)
0.31(0.26‐0.37)
1.19(0.93‐1.51)
3.10(2.43‐3.96)
7.39(6.26‐8.72)
1.60(1.07‐2.39)
1.40(1.30‐1.50)
2.05(1.30‐3.23)
59(25)
0.93(0.36)
1.10(0.27)
Abbreviations:IL‐6,interleukin‐6;PAI‐1,plasminogenactivatorinhibitor‐1;OPG,osteoprotegerin;MCP‐
1,monocytechemotacticprotein‐1;MIF,macrophagemigrationinhibitingfactor;MIP‐1α,macrophage
inflammatoryprotein‐1α,IL‐1RA,interleukin‐1receptorantagonist;SAA‐1,serumamyloidA1;CXCL10,
C‐X‐C motif chemokine 10; OPG, osteoprotegerin. Data are geometric means and 95% confidence
intervals,exceptindicatedotherwise.Independent‐samplest‐testswereusedtoanalyzethedata,andP
values were interpreted applying false discovery rate correction for multiple testing. aValid measures
abovethelowerlimitofdetectionavailableforn=22patients.bMean(SD).Noneofthecomparisonswere
statisticallysignificant.
87
Supplemental Figure 1. Correlation between adipose tissue macrophages
quantifiedfrom CD68‐stained paraffin sections andadipose tissue CD68 mRNA
expression.
Abbreviations: ATM, adipose tissue macrophage; CLS, crown‐like structure. A) Correlation between
solitary ATMsandadiposetissueCD68 mRNAexpressioninCOPDpatients(blackcircles)andhealthy
controls (open circles). These correlations were not statistically significant (P>.05). B) Correlation
betweenCLSdensityandadiposetissueCD68mRNAexpressioninsubjectsinwhomweencountered≥1
CLS. The CLS density was significantly correlated with adipose tissue CD68 mRNA expression
(correlationforentiregroupbecauseofthesmallsamplesize:Spearman’sρ=0.53,P=.024,n=17).This
positive correlation remained significant after excluding the outlier at CLS density=1.7 (Spearman’s
ρ=0.50,P=.047,n=16).
88
CHAPTER5
The influence of abdominal visceral
fat on inflammatory pathways and
mortality risk in obstructive lung
disease
BramvandenBorst,HarryR.Gosker,AnnemarieKoster,BinbingYu,
Stephen B. Kritchevsky, Yongmei Liu, Bernd Meibohm, Thomas B.
Rice, Michael Shlipak, Sachin Yende, Tamara B. Harris, Annemie
M.W.J.Schols
FortheHealth,AgingandBodyComposition(HealthABC)Study
89
EXCESSIVEABDOMINALVISCERALFATINCOPD
Abstract
Background: Low‐grade systemic inflammation, particularly elevated IL‐6,
predicts mortality in chronic obstructive pulmonary disease (COPD). Although
alteredbodycomposition,especiallyincreasedvisceralfat(VF)mass,couldbea
significant contributor to low‐grade systemic inflammation, this remains
unexploredinCOPD.
Objective: To investigate COPD‐specific effects on VF, plasma adipocytokines
andtheirpredictivevalueformortality.
Design:WithintheHealth,AgingandBodyCompositionStudy,anobservational
studyofcommunity‐dwellingolderpersons,weusedpropensityscorestomatch
n=729 persons with normal lung function to n=243 persons with obstructive
lung disease (OLD, defined as FEV1/FVC<LLN). Matching was based on age,
gender,race,clinicsite,bodymassindexandsmoking.Withinthiswell‐balanced
match, we compared CT‐acquired VF area (VFA) and plasma adipocytokines,
analyzed independent associations of VFA and OLD status on plasma
adipocytokines,andstudiedtheirpredictivevaluefor9.4‐yearmortality.
Results: Whereas whole‐body fat mass was comparable between the groups,
OLD persons had increased VFA and higher plasma IL‐6, adiponectin and
plasminogen activator inhibitor (PAI)‐1. Both OLD status and VFA were
independently positively associated with IL‐6. Adiponectin was positively
associated with OLD status, but negatively with VFA. PAI‐1 was no longer
associatedwithOLDstatusafteraccountingforVFA.OLDpersonshadincreased
risk of all‐cause, respiratory and cardiovascular mortality, of which IL‐6 was
identifiedasanindependentpredictor.
Conclusions:Ourdatasuggestthatexcessiveabdominalvisceralfatcontributes
toincreasedplasmaIL‐6which,inturn,isstronglyassociatedwithall‐causeand
cause‐specificmortalityofolderpersonswithOLD.
90
CHAPTER5
Introduction
Inolderadults,theprevalenceofchronicobstructivepulmonarydisease(COPD)
is 14% and as a consequence of general aging in the population and improved
medical intervention, this prevalence is projected to increase even further [1].
COPD is characterized by altered body composition towards a relative or
absolute increase in fat mass, low‐grade systemic inflammation and high
mortality[2].Classically,low‐gradesystemicinflammationinthesepatientshas
beenconsideredtobetheresultofa‘spill‐over’ofinflammatorymediatorsfrom
the inflamed pulmonary compartment. For this, however, there has been no
convincing evidence during clinically stable disease. The possibility of extra‐
pulmonarytissuescontributingtolow‐gradesystemicinflammationinCOPDhas
beenrelativelyunexplored.
Adipose tissue has emerged as a potent producer of mediators of
inflammation and energy homeostasis, termed adipocytokines [3], including
interleukin (IL)‐6, plasminogen activator inhibitor (PAI)‐1, leptin and
adiponectin. Notably, in diseases associated with excessive fat mass, such as
obesityandtype2diabetes,adiposetissuehasbeensuggestedtobethesource
oflow‐gradesystemicinflammation[4].Inarecentclinicalstudy,weexamined
adipocytokine gene expression and macrophage markers in biopsies from the
abdominal subcutaneous fat (SF) compartment from COPD patients and fat
mass‐matched healthy subjects [5]. We foundpositive associationsbetween fat
massandadiposetissueinflammation,whichwerehighlycomparablebetween
COPDpatientsandcontrols.ThesedatasuggestedthattheSFcompartmentmay
not be the primary fat compartment contributing to low‐grade systemic
inflammation in COPD. Importantly, studies consistently indicate that the
inflammatory capacity of abdominal visceral fat (VF) is considerably greater in
comparison with other fat depots including SF [6, 7]. Thus, VF may be a more
plausiblesourceofsystemicinflammationthanSF.Interestingly,arecentstudy
in normal‐weight mild‐to‐moderate COPD patients and healthy subjects used
abdominalCT‐scanningandfoundincreasedVFarea(VFA)inCOPDpatients[8].
However, these patients also had increased whole‐body fat mass, and it is
unclearwhethertheincreasedVFwasareflectionofthehigherwhole‐bodyfat
mass. Also, associations between VF and circulating adipocytokines were not
assessed.
Whole‐body fat mass increases with aging, and consistent data has shown
thatinthisprocess,theexpansionrateofVFexceedsthatofSF[9,10].Recently,
inareviewofinflammatorymarkersinpopulationstudiesofaging,ithasbeen
proposed that aging‐related mortality is associated with VF accumulation and
‘inflammaging’ [11]. Collectively, these studies stress the need for investigating
the role of VF in low‐grade systemic inflammation and mortality in COPD,
accountingforimportantconfoundersincludingageandbodymassindex(BMI).
The current study provides a comprehensive comparison of VF mass, plasma
91
adipocytokines andtheir relation with 9.4‐year mortality of COPD patients and
propensity‐score matched persons with normal lung function who participated
in the Health, Aging and Body Composition (ABC) Study. Additionally, we
exploreddietaryintakeandphysicalactivityasimportantlifestyledeterminants
inthesesubjects.
Methods
Studypopulation
The observational Health ABC Study included 3075 community‐dwelling black
andwhitemenandwomenbetweentheagesof70‐79yearsresidinginandnear
Pittsburgh,PennsylvaniaandMemphis,Tennessee.Baselinedatawereobtained
in 1997/1998 through in‐person interview and clinic based examination.
Inclusion criteria were no self‐reported difficulty walking a quarter mile,
climbing 10 steps without resting, or performing mobility‐related activities of
dailyliving.Exclusioncriteriawereanylife‐threateningcondition,participation
inanyresearchstudyinvolvingmedicationsormodificationofeatingorexercise
habits,planstomovefromthegeographicareawithin3years,anddifficultyin
communicating with the study personnel or cognitive impairment. The study
was approved by the Institutional Review Boards of the participating centers
University of Tennessee Health Science Center, Memphis (#95‐05531‐FB) and
the University of Pittsburgh (#960212), and of the coordinating center at the
University of California, San Francisco (#10‐03322). Written informed consent
was obtained for all participants. Clinic site, age and race (black/white) were
based on self‐report. In the current study we retrospectively analyzed baseline
phenotypicaldataand9.4‐yearmortalitydata.
Lungfunctionandsmokinghistory
Pre‐bronchodilator lung function was assessed according to international
standards as previously reported [12]. Since post‐bronchodilator lung function
was lacking, participants with a ratio between forced expiratory volume in 1s
(FEV1) and forced vital capacity (FVC) below the age, sex, and race‐normalized
lower limit of normal (LLN) [12‐14] were regarded as having obstructive lung
disease (OLD). This method is exactly the same as applied in previous
publications from the Health ABC Study [12, 15]. Participants with restrictive
lung disease (FEV1/FVC≥LLN but FVC<LLN) were excluded (Figure 1). Normal
lung function was defined as FEV1/FVC≥LLN and FVC≥LLN. Participants with
missing lung function and those with non‐interpretable lung function
measurements according to international criteria were excluded. Cigarette
smoking status (current/former/never) and the number of pack years smoked
(1packyear=20cigarettes/dayfor1year)wereobtainedbasedonself‐report.
92
CHAPTER5
Figure1.Flowdiagramofselectionofstudyparticipants.
Abbreviations:ATS,AmericanThoracicSociety;FEV1,forcedexpiratoryvolumein1s;FVC,forcedvital
capacity;LLN,lowerlimitofnormal;OLD,obstructivelungdisease.
Anthropometryandbodycomposition
Height was measured using a wall mounted stadiometer. Body weight was
assessedtothenearest0.1kgusingastandardbalancebeamscaleandBMIwas
calculatedasweight/height2(kg/m2).WholebodyDXA(Hologic4500Asoftware
version8.21,Waltham,MA,USA)wasusedtoassesstotalfatandfatfreemasses.
Abdominalcomputerizedtomography(CT)
At120kVpand200‐250 mAa10mmCT‐scanoftheabdomenwasacquiredat
the L4‐L5 level. Subjects were placed in the supine position with their arms
above their head and legs elevated with a cushion to reduce the curve in the
back. In Memphis the scan was acquired using a Somatom Plus 4 (Siemens,
Erlangen,Germany)oraPickerPQ2000S(MarconiMedicalSystems,Cleveland,
Oh, USA), and in Pittsburgh using a 9800 Advantage (General Electric,
Milwaukee,WI,USA).VisceralFatArea(VFA)wasmanuallydistinguishedfrom
subcutaneous fat area (SFA) by tracing along the fascial plane defining the
internal abdominal wall. Areas were calculated by multiplying the number of
pixelsofagiventissuebythepixelarea(seeforfurtherdetails[16]).
93
Circulatingadipocytokines
Interleukin(IL)‐6,plasminogenactivatorinhibitor(PAI)‐1,tumornecrosisfactor
(TNF)‐α, C‐reactive protein (CRP), leptin and adiponectin were obtained from
frozenstoredplasmaorserumobtainedfromavenipunctureafteranovernight
fast (a detailed description on measurement techniques was previously
published[17]).
Insulinresistance
The Homeostatic Model Assessment was used to estimate insulin resistance
(HOMA‐IR)from(fastingplasmaglucose*fastingplasmainsulin)/22,5[18].
Prevalentdiseases
Themetabolicsyndromewasdefinedaccordingtointernationalguidelines[19]
asmeeting at leastthree of thefollowing criteria: 1) waist circumference ≥102
cminmenand≥88cminwomen,2)diastolicbloodpressure≥85mmHgand/or
systolic blood pressure ≥130 mmHg or using antihypertensive medications, 3)
fasting glucose level ≥100 mg/dl or using antidiabetic medication, 4) high‐
densitylipoproteincholesterol(HDL)level<40mg/dlinmenand<50mg/dlin
women or currently on treatment for low HDL cholesterol, and 5) serum
triglyceride level ≥150 mg/dl or currently on drug treatment for high
triglycerides.Diabeteswasdefinedbyself‐reportoruseofdiabetesmedication.
Cardiovascular disease was defined by self‐report or by medical records of
coronaryheartdiseaseand/orstroke.Twoseatedrestingbloodpressureswere
taken and were averaged. Physiological hypertension was defined by systolic
bloodpressure>140mmHgand/ordiastolicbloodpressure>90mmHgoruseof
hypertensionmedication.
Survivaltimeandcauseofdeath
Surveillance for survival was conducted by in‐person visit alternating with
telephone interview every 6 months. The dateofdeath was determined on the
basisofhospitalrecords,deathcertificatesorinformantinterviews.Thirty‐eight
persons(3%)werelostduringfollow‐up,andtheirsurvivaltimewascensored
on the basis of their date of last contact. Deaths were adjudicated by a central
committee for immediate and underlying causes of death using established
criteria, including review of death certificate, all recent hospital records, and
interview with the next of kin. Respiratory mortality was defined as mortality
from COPD, pneumonia and respiratory failure. We defined cardiovascular
mortalityasmortalityfromatheroscleroticcardiovasculardisease(definitefatal
myocardial infarction or definite or possible fatal coronary heart disease),
94
CHAPTER5
stroke,atheroscleroticdiseaseotherthancoronaryorcardiovascular,andother
cardiovasculardisease.
Dietaryintakeandphysicalactivitylevel
Usualnutrientandfoodgroupintakewasestimatedbyadministeringamodified
Blockfoodfrequencyquestionnairebyatraineddietaryintervieweratthefirst
annualfollow‐upexamination(detailspublishedin[20]).AHealthyEatingIndex
(HEI) was calculated to measure the amount of variety in the diet and
compliance with specific dietary guidelines [20]. HEI scores ranged from 0 to
100 with higher scores indicating bettercompliance with recommended intake
rangeoramount.Physicalactivitylevelwasassessedatbaselinebymeansofa
validatedquestionnaire[20].
Statisticalanalyses
Propensityscorematching
Afterapplyingexclusioncriteria,n=2139participants(n=243OLDpersonsand
n=1896non‐OLDpersons)wereidentified(Figure1).Bymeansofindependent
t‐tests, Mann‐Whitney‐U tests and χ2‐tests as appropriate, sex, age, body mass
index(BMI),packyearsandsmokingstatusweresignificantlydifferentbetween
OLD persons and non‐OLD persons(Table 1). To enable balanced comparisons
between OLD persons and non‐OLD persons accounting for these confounders,
we performed propensity score matching [21]. For this, propensity scores for
OLD status were calculated for the entire population (n=2139) using logistic
linear regression based on sex, age, race, clinic site, BMI, smoking status and
packyears.Subsequently,threenon‐OLDparticipantswerematchedtoeachOLD
person with the closest propensity score. In this procedure we allowed for
replacement of non‐OLD persons, which has been shown to increase balance
[22].
Fourmethodswereemployedtoassessthesuccessofthematching.First,we
confirmedthatnoneofthevariablesincludedinthepropensityscorecalculation
werestatisticallydifferentanymorebetweenOLDpersonsandnon‐OLDpersons
after matching (Table 1), indicating that a balanced match was reached. Also,
comparisonsofthemeanpropensityscoresbetweenOLDpersonsandnon‐OLD
persons before (0.211±0.149 vs 0.101±0.093, P<.001) and after matching
(0.211±0.149 vs 0.211±0.149, P=.99) showed a perfect match. Additionally, the
standardizeddifferencesbetweentheOLDpersonsandnon‐OLDpersonsofthe
variablesincludedinthepropensityscorecalculationwerecomparedbeforeand
after matching (Table 1). The standardized differences were calculated as the
absolute difference in sample means divided by the standard deviation of the
total population, expressed as a percentage [23]. The standardized differences
95
wereconsiderablyimprovedbythematching,reachingacceptablelevels.Finally,
empirical‐QQ plots were created before and after matching for the continuous
variables included in the propensity score calculation (i.e., age, BMI and pack
years).TheseplotsallowforavisualinspectionofthedatadistributioninOLD
persons and non‐OLD persons and show that the equality of data distribution
wasmarkedlyimprovedbythematchingprocedure(Figure2).
Table1.Sociodemographicdatabeforeandafterpropensityscorematching.
Beforematching
Aftermatching
Non‐OLD Standardized
OLD
Non‐OLD Standardized
persons
(n=243)
persons
(n=1896)
Sex,%men
58
47b
Age,y
73(3)
74(3)c
Race,%white
54
61
Clinicsite,
51
46
%Memphis
25.6(4.6) 27.4(4.5)b
BMI,kg/m2
Packyearsd
38(9‐57) 2(0‐25)b
Smokingstatus
Current,%
27.6
7.8
Former,%
54.7
44.5b
Never,%
17.7
47.7
difference
(%)
persons difference(%)
(n=729)a
22.8
19.9
13.9
9.2
58
73(3)
56
48
0.0
0.2
2.8
1.5
40.6
84.5
25.7(4.1)
30(6‐63)
66.0
20.5
60.4
18.6
61.6
19.9
4.1
3.3
22.4
14.1
5.6
Abbreviations: BMI, body mass index; OLD, obstructive lung disease. Data are mean (SD) unless
indicated otherwise. aNone of these variables were statistically different compared to OLD persons as
testedwithrandom‐effectsanalysisofvariance. bP<.001, cP<.01comparedtoOLDpersonsastestedby
independentt‐tests,Mann‐Whitney‐UtestsandΧ2‐testsasappropriate.dMedian(IQR).
Analyseswithintheestablishedmatch
Itiscrucialthatcomparisonsdoneafterpropensityscorematchingaccountfor
the lack of independence between matched sets [23]. We used random‐effects
analysis of variance to test phenotypical differences between OLD persons and
matched non‐OLD persons. Association analyses were performed using linear
mixed models in which matched sets were treated as random factors. Cox
proportionalhazardmodelsaccountingforcorrelateddata[24]wereperformed
toinvestigatethehazardratio(HR)ofall‐cause,respiratoryandcardiovascular
mortality of OLD persons relative to matched non‐OLD persons. These models
were first performed unadjustedand adjusted for thevariables included in the
propensityscorecalculation.Subsequently,thephenotypicalvariablesthatwere
significantly different between OLD persons and matched non‐OLD persons
96
CHAPTER5
Figure 2. Empirical‐QQ plots for age (A), body mass index (B), and pack years
(C)beforeandaftermatching.
Quantilesofage,bodymassindex(BMI)andpackyearsofobstructivelungdisease(OLD)personsand
non‐OLD persons were plotted against each other before and after matching. The solid line indicates
y=x,whichrepresentsthesituationinwhichthetwodistributionsbeingcomparedareequal.Itcanbe
observed that the matching improved balance considerably. The “before matching” figures describe
datafromn=243OLDpersonsandn=1896non‐OLDpersons,andthe“aftermatching”figuresdescribe
datafromn=243OLDpersonsandn=729non‐OLDpersons.
wereaddedtothemodeltotestwhetherthesepredictedmortalityandmodified
mortalitybyOLDstatus.
97
R version 2.11.1 was used to establish the match (Matching package), to
perform the linear mixed models (nlme package), and to perform the
proportional hazard models (Survival package). Random‐effects analysis of
variance tests were performed in PASW Statistics 17.0 (SPSS inc. Chicago, IL,
USA).DifferencesweredeclaredtobestatisticallysignificantatP<.05.
Results
PhenotypicaldifferencesbetweenOLDpersonsandmatchednon‐OLDpersons
Table 2 summarizes the main phenotypical characteristics of OLD persons and
matched non‐OLD persons. On the whole‐body level, fat and fat‐free masses
werenotdifferent.Yet,OLDpersonshadsignificantlygreaterVFA(P<.001)but
nodifferentSFA(Figure3).CirculatinglevelsofIL‐6,adiponectinandPAI‐1were
significantlyhigherinOLDpersonswhereaslevelsofCRP,TNF‐αandleptinwere
notdifferent.ApartfromahigherprevalenceofhypertensioninOLDpersons,no
differences were found for HOMA‐IR, diabetes, cardiovascular disease and
metabolicsyndromecriteria.
Associationanalyses
Our findings of concomitantly increased VFA, IL‐6, adiponectin and PAI‐1 raise
the suggestion that VF may be a significant contributor to these elevated
circulating adipocytokines. Table 3 summarizes the data from the linear mixed
effectsmodelsstudyingtheindependentassociationsofVFAandOLDstatuson
these plasma adipocytokine concentrations. VFA and OLD status were both
significantlyand independently associated with IL‐6. OLD status was positively
associated with adiponectin, independent of VFA, and VFA associated negative
with adiponectin. OLD status was no longer associated with PAI‐1 after
accountingforVFAwhichwassignificantlyassociatedwithPAI‐1.Thedatafrom
thesemodelsdidnotchangeafterfurtheradjustmentforBMI,suggestingspecific
VFeffects.
Mortalityanalyses
During a median follow‐up period of 9.4 years (7728 person years), 104 OLD
persons (43%) and 201 non‐OLD persons (28%) died from all causes.
Unadjusted Cox proportional hazards models revealed a significantly worse
survival in OLD persons compared to matched non‐OLD persons for all‐cause
mortality (HR 1.70 [95% CI 1.29, 2.34], P<.001) (Figure 4A). As expected,
respiratory mortality risk was significantly higher in OLD persons (HR 4.03
[95%CI2.12,7.64],P<.001)(Figure4B).Alsocardiovascularmortalityriskwas
98
CHAPTER5
Table 2. Phenotypical characteristics of OLD persons and matched non‐OLD
persons.
OLDpersons(n=243)
Non‐OLDpersons(n=729)
FEV1,%preda
63(18)
99(16)
<.001
FVC,%preda
81(18)
98(14)
<.001
FEV1/FVC,%
57(7)
75(5)
<.001
Pulmonaryfunction
Bodycomposition
Weight,kg
P
72.6(14.9)
72.3(14.4)
.76
Fatmass,kg
23.9(8.4)
23.4(7.2)
.33
Fatfreemass,kg
48.7(10.5)
49.0(10.7)
.72
Subcutaneousfatarea,cm2
246(118)
238(101)
.31
Visceralfatarea,cm2
143(77)
123(59)
<.001
Systemicadipocytokines
IL‐6,pg/mlb
CRP,µg/mlb
2.16(1.52‐3.34)
1.75(1.20‐2.69)
<.001
2.0(1.2‐3.7)
1.4(0.9‐2.6)
.12
TNF‐α,pg/mlb
3.03(2.35‐4.08)
3.10(2.40‐4.07)
.77
PAI‐1,ng/mlb
22(12‐37)
18(11‐31)
.008
Adiponectin,µg/ml
12.3(7.4)
11.3(6.7)
.037
7.61(3.33‐14.07)
7.86(4.22‐14.28)
.91
.33
Leptin,ng/mlb
Insulinsensitivity
HOMA‐IR 1.5(1.1‐2.3)
1.5(1.0‐2.2)
Metabolicsyndrome
Abdominalobesity,%
48
45
.42
.57
b
Highbloodpressure,%
79
77
Highglucose,%
21
20
.86
LowHDL,%
24
26
.36
Hightriglycerides,%
21
20
.14
Prevalentdiseases
Hypertension,%
66
56
.010
Diabetes,%
12
13
.71
Cardiovasculardisease,%
22
28
.27
Abbreviations: FEV1, forced expiratory volume in 1s; FVC, forced vital capacity; CRP, C‐reactive
protein;TNF‐α,tumornecrosisfactor‐α.HOMA‐IR,homeostaticmodelassessmentinsulinresistance;
HDL, high density lipoprotein. Data are mean (SD) unless indicated otherwise. aFrom reference
equations[13].bMedian(IQR).Analyseswereperformedusingrandom‐effectsANOVA.
significantlyhigherinOLDpersons(HR1.76[95%CI1.07,2.89],P=.026)(Figure
4C).
Table4describestheanalysesofpredictorsformortality.Adjustmentforthe
variablesincludedinthepropensityscorematchinghadnomajoreffectonthese
hazard ratios (Model 2). Because we found increased VFA, elevated IL‐6,
adiponectin and PAI‐1, and higher hypertension prevalence in OLD persons
compared to matched non‐OLD persons, we studied whether these predicted
99
Figure 3. Excessive visceral fat accumulation but normal subcutaneous fat
contentinOLD.
Abdominal CT‐scans from a healthy person (A) and a COPD patient (FEV1 53% of predicted) (B)
matchedforage,sex,andbodymassindex(29kg/m2).TheblackandwhiteareasinCandDdenotethe
visceralandsubcutaneousfatcompartments,respectively.
mortality (Model 3). IL‐6 and adiponectin independently predicted all‐cause
mortality.Inthefinalmodel,onlyIL‐6remainedsignificant.IL‐6wasalsofound
to be a strong and independent predictor of respiratory and cardiovascular
mortality.PAI‐1,VFAandhypertensionwerenopredictorsofmortality.
Table5summarizestheprimaryandunderlyingcausesofmortalityinOLD
personsandmatchednon‐OLDpersons.Primaryrespiratoryandcardiovascular
causes were particularly common in OLD persons, whereas cardiovascular
causes were most represented among non‐OLD persons. Respiratory causes
werereportedastheunderlyingcauseofmortalityin15%ofOLDpersonsandin
only 3% in non‐OLD persons, whereas cardiovascular causes underlying
mortalityweremostcommoninnon‐OLDpersons.
100
CHAPTER5
Figure4.All‐cause(A),respiratory(B),andcardiovascular(C)mortalityofOLD
personsandmatchednon‐OLDpersons.
Abbreviations:OLD,obstructivelungdisease.PlotsarefromunadjustedCoxproportionalhazardmodels
fromn=972persons.
Dietaryintakeandphysicalactivitylevel
Asexplained,packyearssmokedandsmokingstatuswerematchedbetweenthe
OLDpersonsandnon‐OLDpersons.Consequently,thecurrentstudypopulation
consisted of mainly current/former smokers with a relatively high number of
packyears.
Food Frequency Questionnaire data were available of n=210 OLD persons
andn=565non‐OLDpersons(Table6).Whereastotalcalories,totalproteinand
total carbohydrate intake were not different between the groups, OLD persons
had significantly higher intakes of total fat, saturated fat, cholesterol and trans
fat.Also,OLDpersonshadsignificantlylowerintakeofdietaryfibers,whichwas
attributabletoaloweramountoffibersfromfruits/vegetables(datanotshown).
Additionally, daily vitamin C intake was significantly lower in OLD persons.
Whereas daily glycemic load was not different, daily glycemic index was
significantly higher in OLD persons. On average, HEI scores were within the
rangethathasbeenclassifiedas‘needsimprovement’inbothOLDpersonsand
non‐OLDpersons,butwereevenlowerinOLDpersons.
Total physical activity was significantly lower in OLD persons compared to
non‐OLDpersons(Table6).Morespecifically,fewerOLDpersonswalkedbriskly
for ≥90min/wk and fewer OLD persons performed high‐intensity exercise for
≥90min/wkthannon‐OLDpersons.
Discussion
WeaimedtounravelOLD‐specificeffectsfromage‐relatedeffectsonVFmass,to
study the associations between VF and adipocytokines by OLD status and to
investigate their relation with mortality. We found that OLD persons had
significantly increased VF mass independent of age, BMI and whole‐body fat
101
Table 3. Mixed model analyses investigating the independent associations of
OLDstatusandVFAonplasmaconcentrationsofIL‐6,adiponectinandPAI‐1.
Model1
Model2
Model3
β(SE)
P
β(SE)
P
β(SE)
P
IL‐6(logtransformed)
OLDstatus
Non‐OLD
0[Ref]
OLD
0.10 (0.02)
VFA,dm2
Adiponectin(μg/ml)
OLDstatus
Non‐OLD
0[Ref]
OLD
1.04(0.45)
VFA,dm2
PAI‐1(logtransformed)
OLDstatus
Non‐OLD
0[Ref]
OLD
0.07 (0.02)
VFA,dm2
<.001
.022
.004
0[Ref]
0.08(0.02)
0.06(0.02)
0[Ref]
<.001 0.08(0.02)
<.001 0.07(0.02)
0[Ref]
0[Ref]
1.93(0.42) <.001 1.69(0.43)
‐4.39(0.30) <.001 ‐3.37(0.40)
0[Ref]
0.02 (0.02)
0.22(0.02)
0[Ref]
.87
0.03 (0.02)
<.001 0.20 (0.02)
<.001
<.001
<.001
<.001
.69
<.001
Abbreviations:IL‐6,interleukin‐6;OLD,obstructivelungdisease;PAI‐1,plasminogenactivatorinhibitor‐
1;Ref,reference; VFA,visceralfatarea.Analyseswereperformedusinglinearmixedmodels. Model 1:
Adjustedforage,gender,race,clinicsite,smokingstatusandpackyearssmoked.Model2:Model1plus
adjustmentforVFA.Model3:Model2plusadjustmentforbodymassindex.
mass.OurdatasuggestthatthisexcessiveVFcontributestoincreasedplasmaIL‐
6, which was subsequently shown to be a strong predictor of all‐cause,
respiratoryandcardiovascularmortality.Inaddition,OLDpersonsengagedinan
unhealthierlifestylethatwascharacterizedbypoorerdietaryqualityandlower
dailyphysicalactivitylevel.
Making use of the wealthof data inthe Health ABCStudy, we were able to
carefully match a non‐OLD control group to OLD participants using propensity
score matching for important confounders. This approach allowed us to
disentangle OLD‐specific effects from age‐related effects for body fat
distribution, inflammation and mortality. We found that a number of plasma
adipocytokineswerenotdifferentbetweenOLDandmatchednon‐OLDpersons.
It could be possible that whole‐body fat mass rather than specific VF may be
implicated, and as whole‐body fat mass was matched, no differences in these
markerswereobserved.
102
CHAPTER5
Table 4. Hazard ratios (95% confidence interval) of all‐cause, respiratory, and
cardiovascularmortalityaccordingtoOLDstatusandfactorsaccountingforthe
relationship(per1SD‐increaseforcontinuousvariables).
Model1
All‐causemortality
OLDstatus
Non‐OLD
1[Ref]
OLD
1.70(1.29‐2.34)a
IL‐6
Adiponectin
PAI‐1
VFA
Hypertension
No
Yes
Respiratorymortality
OLDstatus
Non‐OLD
1[Ref]
OLD
4.03(2.12‐7.64)a
IL‐6
Adiponectin
PAI‐1
VFA
Hypertension
No
Yes
Cardiovascularmortality
OLDstatus
Non‐OLD
1[Ref]
OLD
1.76(1.07‐2.89)c
IL‐6
Adiponectin
PAI‐1
VFA
Hypertension
No
Yes
Model2
Model3
Model4
1[Ref]
1[Ref]
1[Ref]
1.72(1.30‐2.26)a 1.44(1.05‐1.98)c 1.52(1.14‐2.03)b
1.44(1.28‐1.61)a 1.41(1.26‐1.58)a
1.23(1.02‐1.48)c 1.18(0.98‐1.41)d
1.10(0.96‐1.25)
1.13(0.89‐1.43)
1[Ref]
1.20(0.86‐1.67)
1[Ref]
1[Ref]
1[Ref]
4.13(2.19‐7.80)a 3.61(1.80‐7.25)a 3.66(1.94‐6.91)a
1.50(1.21‐1.83)a 1.47(1.18‐1.84)a
1.21(0.80‐1.82)
1.22(0.82‐1.58)
0.96(0.65‐1.41)
1[Ref]
0.93(0.48‐1.82)
1[Ref]
1[Ref]
1[Ref]
1.75(1.06‐2.89)c 1.49(0.85‐2.60) 1.62(0.96‐2.75)d
1.40(1.14‐1.72)b 1.36(1.11‐1.66)b
1.32(0.99‐1.77)d
1.12(0.87‐1.44)
1.10(0.74‐1.64)
1[Ref]
1.24(0.65‐2.35)
Abbreviations:IL‐6,interleukin‐6;OLD,obstructivelungdisease;PAI‐1,plasminogenactivator‐inhibitor‐
1;Ref,reference;VFA,visceralfatarea.AnalyseswereperformedusingCoxproportionalhazardmodels.
Model1:unadjusted.Model2:Adjustedforage,gender,race,clinicsite,bodymassindex,smokingstatus
and pack years. Model 3: Model 2 plus adjustment for IL‐6, adiponectin, PAI‐1, visceral fat area and
hypertension. Model 4: Model 2 plus adjustment for significant predictors of Model 3. aP<.001. bP<.01.
cP<.05.dP≤.08.
103
Table5.PrimaryandunderlyingcausesofOLDdeathsandnon‐OLDdeaths.
Primarycauseofdeath Underlyingcauseofdeath
OLDdeaths non‐OLD
OLDdeaths non‐OLD
(n=104)
deaths
(n=104)
deaths
(n=201)
(n=201)
Cardiovascular,%a
32
30
33
34
Respiratory,%b
26
11
15
3
Cancer,%
1
0
33
40
Gastrointestinalbleed,
10
9
2
6
renalfailureorsepsis,%
Othercorunknown,%
32
49
16
17
Cardiovascular causes included atherosclerotic cardiovascular disease (definite fatal myocardial
infarction,definitefatalcoronaryheartdisease,possiblefatalcoronaryheartdisease),cerebrovascular
disease, atherosclerotic disease other than coronary or cerebrovascular, other cardiovascular disease
(including valvular heart disease and other). bRespiratory causes included chronic obstructive
pulmonarydisease,pneumoniaandrespiratoryfailure. cOthercausesincludeddementia,diabetes,and
stillothercauses.
a
Ourresultscannotbegeneralizedtoallolderpersons,especiallythosewith
severe OLD, because participants in the Health ABC study were selected to be
abletowalkaquarterofamileandtoclimbtenstepswithoutresting.Ourdata
apply to an older OLD population with on average mild airflow obstruction. A
limitation in the Health ABC Study is that no post‐bronchodilator pulmonary
functionisavailable,precludingtheapplicationofGlobalinitiativeofObstructive
Lung Disease criteria for defining COPD. However, instead of using the same
cutoffsforallindividuals,weusedstringentcriteriabaseonage‐,sex‐andrace‐
adjustedLLNcutpointtodefineOLDinthisolderandmulti‐ethnicpopulation,
as recommended by prior studies [25], and used in prior Health ABC Study
publications[12,14,15].
Elevated plasma IL‐6 has been consistently reported in COPD patients [26‐
30] and has recently been identified as an important biomarker with added
predictive value for mortality in addition to clinical predictors [31]. It is
estimated that about one‐third of the plasma IL‐6 originates from adipose
tissues,andinobesesubjectsvisceraladipocytesproduce3‐foldmoreIL‐6than
subcutaneousadipocytes[7].Otherpotentialsourcesmayincludea“spill‐over”
from the pulmonary compartment, but the liver, peripheral skeletal muscle,
circulating immune cells and dietary factors have also been suggested to
contribute to systemic inflammation. IL‐6 was previously shown to be an
independent predictor for mortality in various chronic aging‐related diseases
such as chronic kidney disease [32], peripheral artery disease [33], and also in
OLD[14].Importantly,IL‐6alsohasbeenidentifiedasapredictorofmortalityin
older populations, after adjustment for chronic diseases [11, 34]. Our data
strengthen previous studies identifying IL‐6 as an important biomarker of
104
CHAPTER5
Table 6. Dietary intake and physical activity level of OLD and non‐OLD
participants.
OLDpersons
non‐OLD
P
persons
Dietaryintakea
Calories,kcal/d
2021(824)
1963(791)
.36
Protein,g/d
70.8(29.4)
69.1(30.4)
.49
Carbohydrate,g/d
253(109)
254(105)
.91
Totalfat,g/d
81.5(39.5)
74.9(38.7) .034
Saturatedfat,g/d
24.3(12.6)
21.5(11.3) .003
Cholesterol,mg/d
250(154)
216(141)
.003
Transfat,g/d
8.3(6.1)
7.4(5.1)
.036
Totaldietaryfiber,g/d
16.4(6.9)
18.2(8.4)
.007
VitaminC,mg/d
129(71)
144(79)
.014
Dailyglycemicload,glucosescale
135(62)
133(58)
.64
Dailyglycemicindex,glucosescale
56.7(4.3)
56.0(4.7)
.037
Healthyeatingindex
65.0(13.0)
67.2(12.9) .030
Physicalactivitylevelb
c
Totalphysicalactivity,kcal/kg/wk
65(36‐99)
67(41‐110) .039
Walkingbriskly≥90min/wk,%
6.6
17.3
<.001
High‐intensityexercise≥90min/wk,%
8.2
17.6
<.001
Dataaremeans(SD)unlessspecifiedotherwise.Analyseswereperformedusingrandom‐effectsANOVA.
aAvailableforn=210OLDpersonsandn=656non‐OLDpersons. bAvailableforn=243OLDpersonsand
n=729non‐OLDpersons.cMedian(IQR).
mortalityrisk,andextendthisbythefindingthatexcessiveVFisassociatedwith
increased IL‐6 in OLD. Future studies are warranted to examine the
inflammatory status in VF biopsies from COPD patients and BMI‐matched
controlsubjectstofurtherelucidatethecontributionofVFtolow‐gradesystemic
inflammationinCOPD.
Adiponectin is almost exclusively produced by adipocytes and is typically
described as an insulin sensitizer with anti‐inflammatory properties [35].
Paradoxically, increased circulating adiponectin was found to be a strong
independentpredictorofmortalityinageneralolderpopulation[36],butalsoin
variouswasting‐associateddiseasessuchaschronicheartfailure(CHF)[37,38],
chronic kidney disease [39], respiratory failure [40] and COPD [41]. It is
noteworthy that adiponectin circulates in high, middle, and low‐molecular
weight isoforms. These isoforms have different affinities with the adiponectin
receptors and therefore their mode of action may differ as well. We measured
only total adiponectin in this study, but it might be relevant to explore
adiponectin isoforms in future studies because SF and VF release different
isoforms [42]. Interestingly, whereas we found thatVF was strongly negatively
105
associatedwithplasmaadiponectin,OLDpersons‐whohadincreasedVF‐had
elevated adiponectin levels. This suggests that other fat depots or even other
organ systems are implicated in the elevated adiponectin levels in OLD. In a
recent study it was found that adiponectin was highly expressed in pulmonary
epithelium, and this pulmonary adiponectin expression was strongly increased
in emphysematous COPD patients compared to control subjects [43].
Adiponectinmayplayaroleinpulmonaryinflammationafterozoneexposurein
mice[44],thoughitsfunctioninCOPD‐associatedinflammationremainsunclear.
It has been proposed that elevated adiponectin levels increase energy
expenditureandinduceweightlossthroughadirecteffectonthebrain[38,45],
which would be unfavorable in chronic wasting‐associated diseases, including
COPD.Alternatively,elevatedadiponectinlevelsinCOPDandCHF[46]maybea
sign of adiponectin resistance at the level of skeletal muscle. Lower expression
levels of adiponectin receptors in skeletal muscle have been associated with
insulin resistance [47] and, interestingly, Van Berendoncks et al. [48] recently
showed lower adiponectin receptor expression in skeletal muscle biopsies of
CHFpatients.Thesedatasuggestanimportantcross‐talkbetweenadiposetissue
andskeletalmusclethatwarrantsfurtherinvestigationinCOPDpatients.
FuturestudiesarenecessarytoidentifyfactorsassociatedwithincreasedVF
accumulationinCOPD.Poordietaryqualityanddecreasedphysicalactivitylevel
deservefurtherattentionwithrespecttobodyfatdistributioninCOPDbutalso
point towards the need of more and integrated attention to diet and physical
activity in addition to smoking as lifestyle determinants of COPD risk and
progression.Also,hypogonadismhasbeenassociatedwithCOPD[49]andwith
VFaccumulation[50].Interestingly,increasedVFmasshasalsobeenobservedin
other diseases associated with chronic inflammation, including rheumatoid
arthritis[51],Crohn’sdisease[52],andpsoriasis[53],supportingacentralrole
forVFinchronicinflammatorydiseases.
Inconclusion,ourstudydemonstratesincreasedVF(independentoftotalfat
mass) and a possible role of VF in inflammatory pathways associated with
mortalityinolderpersonswithOLD.
106
CHAPTER5
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
GooneratneNS,PatelNP,CorcoranA.Chronicobstructivepulmonarydiseasediagnosisandmanagementinolder
adults.JAmGeriatrSoc.2010;58(6):1153‐62.
Patel AR, Hurst JR. Extrapulmonary comorbidities in chronic obstructive pulmonary disease: state of the art.
ExpertRevRespirMed.2011;5(5):647‐62.
2004;92(3):347‐55.
Wood IS, de Heredia FP, Wang B, Trayhurn P. Cellular hypoxia and adipose tissue dysfunction in obesity. Proc
NutrSoc.2009;68(4):370‐7.
vandenBorstB,GoskerHR,WesselingG,deJagerW,HellwigVA,SnepvangersFJ,ScholsAM.Low‐gradeadipose
tissue inflammation in patients with mild‐to‐moderate chronic obstructive pulmonary disease. Am J Clin Nutr.
2011;94(6):1504‐12.
KoenenTB,StienstraR,vanTitsLJ,JoostenLA,vanVelzenJF,HijmansA,PolJA,vanderVlietJA,NeteaMG,Tack
CJ, et al. The inflammasome and caspase‐1 activation: a new mechanism underlying increased inflammatory
activityinhumanvisceraladiposetissue.Endocrinology.2011;152(10):3769‐78.
Fried SK, Bunkin DA, Greenberg AS. Omental and subcutaneous adipose tissues of obese subjects release
interleukin‐6:depotdifferenceandregulationbyglucocorticoid.JClinEndocrinolMetab.1998;83(3):847‐50.
FurutateR,IshiiT,WakabayashiR,MotegiT,YamadaK,GemmaA,KidaK.Excessivevisceralfataccumulationin
advancedchronicobstructivepulmonarydisease.IntJChronObstructPulmonDis.2011;6:423‐30.
KukJL,LeeS,HeymsfieldSB,RossR.Waistcircumferenceandabdominaladiposetissuedistribution:influenceof
ageandsex.AmJClinNutr.2005;81(6):1330‐4.
SugiharaM,OkaR,SakuraiM,NakamuraK,MoriuchiT,MiyamotoS,TakedaY,YagiK,YamagishiM.Age‐related
changesinabdominalfatdistributioninJapaneseadultsinthegeneralpopulation.InternMed.2011;50(7):679‐85.
SinghT,NewmanAB.Inflammatorymarkersinpopulationstudiesofaging.AgeingResRev.2011;10(3):319‐29.
Yende S, Waterer GW, Tolley EA,Newman AB,Bauer DC,Taaffe DR, JensenR, CrapoR,RubinS,Nevitt M, etal.
Inflammatory markers are associated with ventilatory limitation and muscle dysfunction in obstructive lung
diseaseinwellfunctioningelderlysubjects.Thorax.2006;61(1):10‐6.
Hankinson JL, Odencrantz JR, Fedan KB. Spirometric reference values from a sample of the general U.S.
population.AmJRespirCritCareMed.1999;159(1):179‐87.
MehrotraN,FreireAX,BauerDC,HarrisTB,NewmanAB,KritchevskySB,MeibohmB.Predictorsofmortalityin
elderlysubjectswithobstructiveairwaydisease:thePILEscore.AnnEpidemiol.2010;20(3):223‐32.
vandenBorstB,KosterA,YuB,GoskerHR,MeibohmB,BauerDC,KritchevskySB,LiuY,NewmanAB,HarrisTB,
etal.Isage‐relateddeclineinleanmassandphysicalfunctionacceleratedbyobstructivelungdiseaseorsmoking?
Thorax.2011;66(11):961‐9.
Snijder MB, Visser M, Dekker JM, Seidell JC, Fuerst T, Tylavsky F, Cauley J, Lang T, Nevitt M, Harris TB. The
prediction of visceral fat by dual‐energy X‐ray absorptiometry in the elderly: a comparison with computed
tomographyandanthropometry.IntJObesRelatMetabDisord.2002;26(7):984‐93.
KosterA,StenholmS,AlleyDE,KimLJ,SimonsickEM,KanayaAM,VisserM,HoustonDK,NicklasBJ,TylavskyFA,
etal.BodyFatDistributionandInflammationAmongObeseOlderAdultsWithandWithoutMetabolicSyndrome.
Obesity(SilverSpring).2010;18(12):2354‐61.
Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment:
insulin resistance and beta‐cell function from fasting plasma glucose and insulin concentrations in man.
Diabetologia.1985;28(7):412‐9.
GrundySM,CleemanJI,DanielsSR,DonatoKA,EckelRH,FranklinBA,GordonDJ,KraussRM,SavagePJ,SmithSC,
Jr.,etal.Diagnosisandmanagementofthemetabolicsyndrome:anAmericanHeartAssociation/NationalHeart,
Lung,andBloodInstituteScientificStatement.Circulation.2005;112(17):2735‐52.
Lee JS, Kritchevsky SB, Tylavsky FA, Harris T, Everhart J, Simonsick EM, Rubin SM, Newman AB. Weight‐loss
intention in the well‐functioning, community‐dwelling elderly: associations with diet quality, physical activity,
andweightchange.AmJClinNutr.2004;80(2):466‐74.
Sekhon JS. Multivariate and propensity score matching software with automated balance optimization: the
matchingpackageforR.JournalofStatisticalSoftware.2011;42(7).
Ming K, Rosenbaum PR. Substantial gains in bias reduction from matching with a variable number of controls.
Biometrics.2000;56(1):118‐24.
AustinPC.Acriticalappraisalofpropensity‐scorematchinginthemedicalliteraturebetween1996and2003.Stat
Med.2008;27(12):2037‐49.
WeiLJ,LinDY,WeissfeldL.Regressionanalysisofmultivariateincompletefailuretimedatabymodelingmarginal
distributions.JournaloftheAmericanStatisticalAssociation.1989;84:1065‐73.
Swanney MP, Ruppel G, Enright PL, Pedersen OF, Crapo RO, Miller MR, Jensen RL, Falaschetti E, Schouten JP,
HankinsonJL,etal.UsingthelowerlimitofnormalfortheFEV1/FVCratioreducesthemisclassificationofairway
obstruction.Thorax.2008;63(12):1046‐51.
Walter RE, Wilk JB, Larson MG, Vasan RS, Keaney JF, Jr., Lipinska I, O'Connor GT, Benjamin EJ. Systemic
inflammationandCOPD:theFraminghamHeartStudy.Chest.2008;133(1):19‐25.
BreyerMK,RuttenEP,VernooyJH,SpruitMA,DentenerMA,vanderKallenC,VangreevenbroekMM,WoutersEF.
Genderdifferencesintheadiposesecretomesysteminchronicobstructivepulmonarydisease(COPD):Apivotal
roleofleptin.RespirMed.2011;105(7):1046‐53.
107
28. AttaranD,LariSM,TowhidiM,MaralluHG,AyatollahiH,KhajehdalueeM,GhaneiM,BasiriR.Interleukin‐6and
airflowlimitationinchemicalwarfarepatientswithchronicobstructivepulmonarydisease.IntJChronObstruct
PulmonDis.2010;5:335‐40.
29. HeZ,ChenY,ChenP,WuG,CaiS.LocalinflammationoccursbeforesystemicinflammationinpatientswithCOPD.
Respirology.2010;15(3):478‐84.
30. Garcia‐RioF,MiravitllesM,SorianoJB,MunozL,Duran‐TauleriaE,SanchezG,SobradilloV,AncocheaJ.Systemic
inflammationinchronicobstructivepulmonarydisease:apopulation‐basedstudy.RespirRes.2010;11:63.
31. CelliBR,LocantoreN,YatesJ,Tal‐SingerR,MillerBE,BakkeP,CalverleyP,CoxsonH,CrimC,EdwardsLD,etal.
Inflammatorybiomarkersimproveclinicalpredictionofmortalityinchronicobstructivepulmonarydisease.AmJ
RespirCritCareMed.2012;185(10):1065‐72.
32. Barreto DV, Barreto FC, Liabeuf S, Temmar M, Lemke HD, Tribouilloy C, Choukroun G, Vanholder R, Massy ZA.
Plasma interleukin‐6 is independently associated with mortality in both hemodialysis and pre‐dialysis patients
withchronickidneydisease.KidneyInt.2010;77(6):550‐6.
33. DepalmaRG,HayesVW,ChowBK,ShamayevaG,MayPE,ZacharskiLR.Ferritinlevels,inflammatorybiomarkers,
andmortalityinperipheralarterialdisease:asubstudyoftheIron(Fe)andAtherosclerosisStudy(FeAST)Trial.J
VascSurg.2010;51(6):1498‐503.
34. Harris TB, Ferrucci L, Tracy RP, Corti MC, Wacholder S, Ettinger WH, Jr., Heimovitz H, Cohen HJ, Wallace R.
Associations of elevated interleukin‐6 and C‐reactive protein levels with mortality in the elderly. Am J Med.
1999;106(5):506‐12.
35. Ouchi N, Parker JL, Lugus JJ, Walsh K. Adipokines in inflammation and metabolic disease. Nat Rev Immunol.
2011;11(2):85‐97.
36. Poehls J, Wassel CL, Harris TB, Havel PJ, Swarbrick MM, Cummings SR, Newman AB, Satterfield S, Kanaya AM.
Association of adiponectin with mortality in older adults: the Health, Aging, and Body Composition Study.
Diabetologia.2009;52(4):591‐5.
37. Tsutamoto T, Tanaka T, Sakai H, Ishikawa C, Fujii M, Yamamoto T, Horie M. Total and high molecular weight
adiponectin,haemodynamics,andmortalityinpatientswithchronicheartfailure.EurHeartJ.2007;28(14):1723‐
30.
38. KistorpC,FaberJ,GalatiusS,GustafssonF,FrystykJ,FlyvbjergA,HildebrandtP.Plasmaadiponectin,bodymass
index,andmortalityinpatientswithchronicheartfailure.Circulation.2005;112(12):1756‐62.
39. Menon V, Li L, Wang X, Greene T, Balakrishnan V, Madero M, Pereira AA, Beck GJ, Kusek JW, Collins AJ, et al.
Adiponectinandmortalityinpatientswithchronickidneydisease.JAmSocNephrol.2006;17(9):2599‐606.
40. WalkeyAJ,RiceTW,KonterJ,OuchiN,ShibataR,WalshK,deBoisblancBP,SummerR.Plasmaadiponectinand
mortalityincriticallyillsubjectswithacuterespiratoryfailure.CritCareMed.2010;38(12):2329‐34.
41. Waschki B, Kirsten A, Holz O, Muller KC, Meyer T, Watz H, Magnussen H. Physical activity is the strongest
predictor of all‐cause mortality in patients with chronic obstructive pulmonary disease: a prospective cohort
study.Chest.2011;140(2):331‐42.
42. Kovacova Z, Tencerova M, Roussel B, Wedellova Z, Rossmeislova L, Langin D, Polak J, Stich V. The impact of
obesityonsecretionofadiponectinmultimericisoformsdiffersinvisceralandsubcutaneousadiposetissue.IntJ
Obes(Lond).2011Dec6[Epubaheadofprint].
43. MillerM,ChoJY,PhamA,RamsdellJ,BroideDH.Adiponectinandfunctionaladiponectinreceptor1areexpressed
byairwayepithelialcellsinchronicobstructivepulmonarydisease.JImmunol.2009;182(1):684‐91.
44. ZhuM,HugC,KasaharaDI,JohnstonRA,WilliamsAS,VerboutNG,SiH,JastrabJ,SrivastavaA,WilliamsES,etal.
Impactofadiponectindeficiencyonpulmonaryresponsestoacuteozoneexposureinmice.AmJRespirCellMol
Biol.2010;43(4):487‐97.
45. QiY,TakahashiN,HilemanSM,PatelHR,BergAH,PajvaniUB,SchererPE,AhimaRS.Adiponectinactsinthebrain
todecreasebodyweight.NatMed.2004;10(5):524‐9.
46. VanBerendoncksAM,ConraadsVM.Functionaladiponectinresistanceandexerciseintoleranceinheartfailure.
CurrHeartFailRep.2011;8(2):113‐22.
47. Civitarese AE, Jenkinson CP, Richardson D, Bajaj M, Cusi K, Kashyap S, Berria R, Belfort R, DeFronzo RA,
Mandarino LJ, et al. Adiponectin receptors gene expression and insulin sensitivity in non‐diabetic Mexican
AmericanswithorwithoutafamilyhistoryofType2diabetes.Diabetologia.2004;47(5):816‐20.
48. VanBerendoncksAM,GarnierA,BeckersP,HoymansVY,PossemiersN,FortinD,MartinetW,VanHoofV,Vrints
CJ, Ventura‐Clapier R, et al. Functional adiponectin resistance at the level of the skeletal muscle in mild to
moderatechronicheartfailure.CircHeartFail.2010;3(2):185‐94.
49. Van Vliet M, Spruit MA, Verleden G, Kasran A, Van Herck E, Pitta F, Bouillon R, Decramer M. Hypogonadism,
quadricepsweakness,andexerciseintoleranceinchronicobstructivepulmonarydisease.AmJRespirCritCare
Med.2005;172(9):1105‐11.
51. GilesJT,AllisonM,BlumenthalRS,PostW,GelberAC,PetriM,TracyR,SzkloM,BathonJM.Abdominaladiposity
in rheumatoid arthritis: association with cardiometabolic risk factors and disease characteristics. Arthritis
Rheum.2010;62(11):3173‐82.
52. KatznelsonL,FairfieldWP,ZeizafounN,SandsBE,PeppercornMA,RosenthalDI,KlibanskiA.Effectsofgrowth
hormone secretion on body composition in patients with Crohn's disease. J Clin Endocrinol Metab.
2003;88(11):5468‐72.
53. BalciA,BalciDD,YondenZ,KorkmazI,YeninJZ,CelikE,OkumusN,EgilmezE.Increasedamountofvisceralfatin
patientswithpsoriasiscontributestometabolicsyndrome.Dermatology.2010;220(1):32‐7.
108
CHAPTER5
CHAPTER6
Characterizationoftheinflammatory
and metabolic profile of adipose
tissue in a mouse model of chronic
hypoxia
BramvandenBorst,AnnemieM.W.J.Schols,ChieldeTheije,AgnesW.
Boots,S.EleonoreKöhler,GijsH.Goossens,HarryR.Gosker
JournalofAppliedPhysiology2013Mar28[Epubaheadofprint]
109
HYPOXIA,ADIPOSETISSUEINFLAMMATIONANDMETABOLISM
Abstract
Rationale: In both obesity and chronic obstructive pulmonary disease (COPD),
altered oxygen tension in the adipose tissue (AT) has been suggested to evoke
ATdysfunction,subsequentlycontributingtometaboliccomplications.Studying
theeffectsofchronichypoxiaonATfunctionwilladdtoourunderstandingofthe
complex pathophysiology of alterations in AT inflammation, metabolism and
mass seen in both obesityand COPD. This study investigatedthe inflammatory
andmetabolicprofileofATafterchronichypoxia.
Methods: Fifty‐two‐week‐old C57Bl/6J mice were exposed to chronic hypoxia
(8% O2) or normoxia for 21 days after which AT and plasma were collected.
Adipocyte size, AT gene expression of inflammatory and metabolic genes, AT
macrophagedensityandcirculatingadipokineconcentrationsweremeasured.
Results: Food intake and body weight decreased upon initiation of hypoxia.
However, whereas food intake normalized after 10 days, lower body weight
persisted. Chronic hypoxia markedly reduced AT mass and adipocyte size. AT
macrophagedensityandexpressionofEmr1,Ccl2,LepandTnfweredecreased,
whereas Serpine1 and Adipoq expression levels were increased after chronic
hypoxia. Concomitantly, chronic hypoxia increased AT expression of regulators
of oxidative metabolism and markers of mitochondrial function and lipolysis.
Circulating IL‐6 and PAI‐1 concentrations were increased and leptin
concentrationwasdecreasedafterchronichypoxia.
Conclusion:Chronichypoxiaisassociatedwithdecreasedratherthanincreased
AT inflammation, and markedly decreased fat mass and adipocyte size.
Furthermore, our data indicate that chronic hypoxia is accompanied by
significant alterations in AT metabolic gene expression, pointing towards an
enhancedATmetabolicrate.
110
CHAPTER6
Introduction
Obesity and chronic obstructive pulmonary disease (COPD) are two major
chronic conditions with high prevalence worldwide [1, 2], and are increasingly
co‐occurring [3, 4]. In obesity, both decreased and increased oxygen tension
within the adipose tissue (AT) have been suggested to contribute to AT
dysfunction and inflammation, which may subsequently increase the risk of
metaboliccomplications[5‐8].ChronichypoxaemiainCOPD,ontheotherhand,
accelerates loss of body and fat masses, in part by increasing the resting
metabolic rate [9‐11]. However, the effects of chronic hypoxia on AT function
remainincompletelyunderstood.StudyingtheeffectsofchronichypoxiaonAT
function is important as it may add to our understanding of the complex
pathophysiologyofalterationsinATinflammation,metabolismandmassseenin
bothobesityandCOPD.
It is now well‐established that AT is an inflammatory and metabolically
active tissue [12]. Hypoxia may induce AT dysfunction and inflammation via
directeffectsoflowPaO2,endoplasmicreticulumstress,oxidativestressorother
mechanisms [8]. In obesity, the combination of hypertrophying adipocytes and
insufficient neovascularisation may lead to local hypoxic tissue areas and to
hypoxia within large adipocytes lying distant from capillaries [13]. Indeed, in
various mouse models of obesity local areas of hypoxia were demonstrated in
the AT [14‐16]. This was accompanied by increased AT expression of hypoxia‐
responsive markers (eg. glucose transporter [GLUT]‐1 [15, 16], vascular
endothelialgrowthfactor[VEGF][16]),alteredadipokinemRNAexpression(eg.
increased leptin and plasminogen activator‐inhibitor [PAI]‐1, decreased
adiponectin[14]),increasedmRNAexpressionofpro‐inflammatorymarkers(eg.
tumornecrosisfactor[TNF]‐α,interleukin[IL]‐6[16])andincreasedexpression
and presence of F4/80+ macrophages in the AT [15, 16]. In order to further
substantiate the potential direct role of hypoxia on adipocyte inflammation,
manyinvitrostudieshavebeencarriedoutinwhicheithermurine(mostly3T3‐
L1cells)orhumanadipocyteswereexposedtohypoxia.Althoughthesestudies
usedexposuretoextreme,non‐physiologicalPaO2(1%O2),themajorityofthese
studies have indeed reported gross hypoxia‐induced changes rendering
adipocytesmorepro‐inflammatoryandalteringtheiradipokineexpressionand
secretion profiles [14, 16‐19]. Importantly, however, conflicting data have also
been reported [5, 20]. It has been demonstrated that chronic hypoxia
substantiallyloweredtheinflammatoryresponseinprimaryhumanadipocytes
[20],andPaO2inabdominalsubcutaneousATwasinverselyassociatedwiththe
expressionofinflammatorymarkersinthisfatdepot[5].
The physiological cellular response to hypoxia includes switching from
oxidativetoanaerobicmetabolism.Invitrostudieshaveindeedshownincreased
expression of various glucose transporters and other markers of glycolytic
metabolism and decreased expression of mitochondrial components and their
111
major upstream regulator peroxisome proliferator‐activated receptor (PPAR)‐γ
coactivator‐1α (PGC‐1α) [17]. This shift would imply increased lactate
production from adipocytes which is indeed observed in obesity[21] and after
hypoxia [22]. Studies investigating the effect of altered oxygen tension on
lipolysishaveyieldedconflictingresults.Bothextremehypoxia(1%O2)[23]and
hyperoxia(95%O2)[24]enhancedlipolysisin3T3‐L1adipocytes.
ChronichypoxaemiaasaconsequenceofCOPDmayalsoaffectATfunction.
Recent studies have detected increased mRNA expression of the pro‐
inflammatorymarkersCD40,MAPKkinase4,andc‐JunNH2‐terminalkinase[25]
and increased expression of nuclear factor‐κB inhibitors [26] in abdominal
subcutaneous AT biopsies from hypoxaemic, underweight COPD patients when
comparedtonormoxaemic,overweightCOPDpatients.
Although the data on the causal role of hypoxia on AT dysfunction and
inflammation appear robust, there are several limitations that call for more
research.WhereasAThypoxiainobesityischronic[8],theinvitrostudiescited
above investigated hypoxia exposures of 2‐24h which rather induced acute
effects. In addition, the level of hypoxia applied in these studies was very
extreme (1% O2) which does not reflect physiological AT oxygen tension in
obesity(3‐11%O2)[6].Third,giventhefactthatimmunecellswithintheATplay
an important role in (co‐) driving AT dysfunction in obesity [27], an in vivo
modelisthepreferredmodeltoallowfortheinteractionofdifferentcelltypes
underhypoxicconditions.Forthesereasonswehavedesignedamodelinwhich
weexposedmiceto21daysofhypoxia(8%O2).Thisallowedustoinvestigate
whethertheacuteinflammatoryandmetabolicalterationsobservedwithacute,
severe hypoxia also persist in a chronic model with more physiologically
relevantlevelsofhypoxia.
Methods
Experimentalprocedure
Fifty‐two‐week‐old male C57BL/6J mice (Charles River Laboratories) were
exposed to ambient air (normoxia, n=8) or chronic hypoxia (n=8) for 21 days.
We studied 52‐week‐old mice, rather than the more commonly studied young
adultbutnotfully‐grown12‐week‐oldmice,toprecludeincreasingbodyweight
(and expanding fat mass) as a confounder. 52‐week‐old mice remain weight
stableundercontrolconditions.Allmicewerehousedinexperimentalchambers
at 21°C with a 12‐hour dark/light cycle. Mice received standard chow (V1534‐
000 ssniff R/M‐H, ssniff Spezialdiäten GmbH, Soest, Germany) and water ad
libitum.UsingtheproOXP110(BioSpherix,NY,USA)system,O2wasreplacedby
N2inastepwisemannertocreatenormobaricoxygenlevelsof12%(day1),10%
(day2),andfinally8%(60.8mmHg)onday3.Thelatteroxygenconcentration
was maintained for the remainder of the experiment. Three to four mice were
112
CHAPTER6
housed per cage. Daily food intake was determined per cage, and mice were
weighed daily. On day 21, mice were anesthetized with isoflurane gas, the
abdominal cavity was opened and aortic blood was collected into a heparin‐
coated1‐mLsyringe(Becton,DickinsonandCompany,Breda,TheNetherlands).
OxygenlevelsandpHweremeasuredimmediatelyusingtheABL510BloodGas
Analyzer(Radiometer,DiamondDiagnostics,MA,USA)andbloodcellcountwas
determined with the Coulter Ac T Diff hematology Analyzer (Beckman Coulter,
Woerden, The Netherlands). Plasma was stored at ‐80°C until further analyses.
Visceral adipose tissue (VAT, epididymal) and subcutaneous adipose tissue
(SCAT, inguinal) pads were collected bilaterally and weighed. One specimen of
each was fixed in 4% formaldehyde, transferred to 70% ethanol and
subsequentlyembeddedinparaffin.Theotherwassnap‐frozeninliquidnitrogen
and stored at ‐80°C for later mRNA analyses. Lower limb skeletal muscles
(gastrocnemius, tibialis anterior, soleus, extensor digitorum longus and
plantaris), liver and spleen were dissected and weighed. Two normoxic mice
were excluded because of infection. The protocol was approved by the
CommitteeforAnimalCareandUseofMaastrichtUniversity(projectno.2009‐
151).
Plasmaadipokineassays
PlasmaadipokineprofilesweredeterminedusingtheLuminexxMAP‐technology
[28]. A Bio‐Plex (Bio‐Rad Laboratories, Veenendaal, the Netherlands) murine
cytokine 6‐plex panel was used to quantify the concentrations of interleukin
(IL)‐1β, IL‐6, keratinocyte‐derived chemokine (KC), monocyte chemotactic
protein (MCP)‐1, tumor necrosis factor (TNF)‐α and chemokine (C‐C motif)
ligand 5 (CCL5). Additionally, three independent single‐plex panels were
executed to assess the concentrations of leptin, adiponectin and plasminogen
activator‐inhibitor (PAI)‐1. All assays were performed according to
manufacturer’sinstructions.DataanalysiswasdonewithaLuminex100IS 2.3
system (Luminex, Austin, TX, USA) using the Bio‐Plex Manager 4.1.1 software
(Bio‐Rad Laboratories, Veenendaal, The Netherlands). The lower limits of
detection(LLOD)ofthecytokinesandadipokinesmeasuredwere0.74pg/mlfor
IL‐6,3.4pg/mlforTNF‐α,3.3pg/mlforIL‐1β,1.9pg/mlforKC,23.7pg/mlfor
CCL5, 2.4 pg/ml for MCP‐1, 0.7 ng/ml for leptin, 6.5 pg/ml for PAI‐1, and 6.5
μg/mlforadiponectin.Cytokineconcentrationsforthemeasurementsbelowthe
LLODwereimputedusingamaximum‐likelihoodestimationprocedure[29,30],
whichwasrequiredfortheIL‐6andKCdata.Itiswell‐establishedthatsimpler
methods such as assuming the value of one‐half of the LLOD or omitting
undetectablesamplesfromanalysesgeneratebiasedestimatesofthemeasures
[31]. Reproducibility statistics based on imputed data, on the other hand, are
approximatelyunbiasedwhenlessthanhalfofthemeasurementsarebelowthe
LLOD[29]aswasthecaseinourstudy.
113
Adipocytesize
VATandSCATparaffinsections(4μm)werehematoxylin‐eosinstainedandnon‐
overlapping fields were photographed at 200× magnification (Nikon Eclipse
E800,NikonInstrumentsEuropeBV,Amsterdam,TheNetherlands).Theareasof
≥400 unique adipocytes per fat pad were measured using computerized
software(LuciaGF,Version4.81,LaboratoryImaging,Prague,CzechRepublic).
RT‐qPCRanalysis
Total RNA from VAT and SCAT was extracted with the Rneasy® Lipid Tissue
Mini Kit (Qiagen, Venlo, The Netherlands) and reverse‐transcribed into cDNA
using the Transcriptor cDNA synthesis kit (Roche Applied Sciences, Mannheim,
Germany). cDNA was amplified using Sensimix SYBR & Fluorescein Kit (GC
biotech B.V., Alphen a/d Rijn, The Netherlands) on a quantitative PCR machine
(Bio‐Rad laboratories B.V., Veenendaal, The Netherlands). Table 1 presents an
overview of the target genes and primer sequences. Expression levels of target
genes were normalized for the expression of three stable housekeeping genes
(Ppia,Axin1andCanx)usingGeNormsoftware(Primerdesign,Southampton,NY,
USA)[32].
Immunohistochemistry
Deparaffinized 4 μm thick sections of VAT and SCAT were incubated with the
antibody AIA31240 (SanBio, Uden, The Netherlands) directed against
macrophages [33]. Macrophages were quantified on microscopic pictures at
200× magnification, and macrophage density was expressed as number of
macrophagesper100adipocytes.Ifcrown‐likestructures(CLS;aggregationsof
singleoffusedmacrophagesaroundasingleadipocyte)wereencountered,their
densitywasexpressedasnumberofCLSper100adipocytes.
Statistics
Differences between mice exposed to chronic hypoxia or to normoxia were
tested using Student’s t‐test or Mann‐Whitney‐U tests as appropriate. Body
weight changes during the experiment were tested using repeated measures
ANOVA.DifferencesintheresponsesbetweenVATandSCATweretestedusing
linearregressionanalysesinwhichtheinteractiontermofcondition(normoxia
vs hypoxia) and adipose depot (VAT vs SCAT) was evaluated for statistical
significance. Correlations were tested using Pearson’s correlation coefficient r.
Data are presented as means±SEM. Analyses were performed using PASW
Statistics 17.0 (SPSS inc. Chicago, IL, USA). A P‐value <.05 was considered
statisticallysignificant.
114
TNF‐α
Leptin
Adiponectin
Serine(orcysteine)peptidaseinhibitor,cladeE,member1(Serpine1)
β3‐AR
Adrenergicreceptor,beta3(Adrb3)
Celldeath‐inducingDNAfragmentationfactor,alphasubunit‐likeeffectorA(Cidea) Cidea
Peroxisomeproliferatoractivatedreceptorgamma(Pparg)
Sterolregulatoryelement‐bindingprotein1c(Srebp1c)
Lipase,hormonesensitive(Lipe)
Patatin‐likephospholipasedomaincontaining2(Pnpla2)
Phosphoenolpyruvatecarboxykinase1,cytosolic(Pck1)
PeptidylprolylisomeraseA(Ppia)
Axin1(Axin1)
Calnexin(Canx)
Lipidmetabolism
Housekeepers
UCP‐1
Uncouplingprotein1(mitochondrial,protoncarrier)(Ucp1)
TFAM
CCGGCAGAGACGGTTAAAAA
Calnexin
Axin‐1
CyclophilinA
PEPCK
ATGL
HSL
SREBP‐1c
PPAR‐γ
CATAAGCTGGGCAAGTGGTA
GAAGAAGGGTCGCATGGCAAA
GGACACCTCAATAATGTTGGCAC
GGGCTTGCGTCCACTTAGTTC
CATAGGGGGCGTCAAACAG
TGGGCTTCACGTTCAGCAAG
GCCGTGTTAAGGAATCTGCTG
TCCACAGTTCGCAACCAGTT
GACGTTCCAGGACCCGAGTC
CCGCATGAACATCTCCCCA
TCATCCTTTGCCTCCTGGAA
CCCAGATGAGGGAAGGGACT
CTTCATCCACGGGGAGACTG
GCCAGGGTTGCACTAAACAT
CCAACCTGCACAAGTTCCCTT
AAGCCCAGGAATGAAGTCCA
TTTTCACAGGGGAGAAATCG
CCGTTTTTCCATCTTCTTCTTTGG
TGAAGGACTCTGGCTTTGTCT
TGCCCGGACTCCGCAA
ATTGGGATCATCTTGCTGGT
GCAGCGACCTATGATTGACAACC
AGTGGATCATTGAGGGAGAGA
GCTCCAAACCAATAGCACTGAAAGG
GCCCCAGGACGCTCGAT
TTCCTCCTTTCACAGAATTATTCCA CCGCCAGTGCCATTATGG
GAGGCCACAGCTGCTGCAGAA
CAACGCCACTCACATCTACGG
CAGAAGGCACTAGGCGTGATG
GATGTGCGAACTGGACACAG
CGGAAGCCCTTTGGTGACTT
TGCTCTTCTGTATCGCCCAGT
AGGCAACCTGCTGGTAATCA
GCGTACCAAGCTGTGCGATG
Cytochromec‐1 GCATTCGGAGGGGTTTCCAG
TranscriptionfactorA,mitochondrial(Tfam)
Cytochromec‐1(Cyc1)
ACCCTGAGAAAGCGCAATGA
CAACAATGAGCCTGCGAACA
Peroxisomeproliferativeactivatedreceptor,gamma,coactivator1beta(Ppargc1b) PGC‐1β
GCCTCCTCATCCTGCCTAA
TGTTCCTCTTAATCCTGCCCA
CAAGCAGTGCCTATCCAGA
CCAAGCCTTATCGGAAATGA
GTAATGAAAGACGGCACACCCAC
ATGGATCCTACCAAACTGGAT
CAGCGCTGAGGTCAATCTGCC
CCTGCTGTTCACAGTTGCC
GGATGTACAGATGGGGGATG
GGACCAGGGCCTACTTCAAAGAAG
TCGCTGGTAGACATCCATGAACT
Reverseprimer(5’‐3’)
AGAACTTCCACTCTCCCACCCAAA AGGGTAGTGCCCATTGTTGAAGGA
TGACCATCGCCCTGGCCT
CTGTACCTCCACCATGCCAAGT
Forwardprimer(5’‐3’)
Oxidativemetabolism Peroxisomeproliferativeactivatedreceptor,gamma,coactivator1alpha(Ppargc1a) PGC‐1α
PAI‐1
IL‐10
IL‐1β
IL‐6
Adiponectin,C1Qandcollagendomaincontaining(Adipoq)
Interleukin6(Il6)
Leptin(Lep)
Tumornecrosisfactor(Tnf)
Adipocytokines
MCP‐1
Chemokine(C‐Cmotif)ligand2(Ccl2)
F4/80
PFKFB3
EGF‐likemodulecontaining,mucin‐like,hormonereceptor‐likesequence1(Emr1)
Macrophages
Interleukin1beta(Il1b)
6‐phosphofructo‐2‐kinase/fructose‐2,6‐biphosphatase3(Pfkfb3)
GLUT‐1
Interleukin10(Il10)
Solutecarrierfamily2(facilitatedglucosetransporter),member1(Slc2a1)
Protein
VEGF‐A
VascularendothelialgrowthfactorA(Vegfa)
Gene
Hypoxia
Table1.Genenames,proteinsencodedbythegenes,andprimersequences.
Category
Results
Chronichypoxiainducesmajorhematologicaladaptations
Table 2 summarizes the blood gas analyses and hematological adaptations to
chronic hypoxia. As expected, arterial PaO2 and SaO2 were markedly lower in
miceexposedtochronichypoxia.Furthermore,miceexposedtochronichypoxia
had lower pH and lower HCO3‐ but normal PaCO2. Further differences included
increased hematocrit, erythrocyte count, hemoglobin concentration, mean
corpuscular volume, mean corpuscular hemoglobin, and a decreased mean
corpuscular hemoglobin concentration in mice exposed to chronic hypoxia. In
linewiththeacidosis,thesignificantlyhigherp50(a)(oxygentensionofbloodat
halfsaturation)inmiceexposedtochronichypoxiaindicatesaright‐shiftofthe
oxygen‐hemoglobin dissociation curve, which means a lower affinity of
hemoglobin for oxygen, suggesting improved delivery of oxygen to the tissues.
Also, chronic hypoxia increased spleen weight by 263% (P<.001, data not
shown).
Table2.Arterialbloodgasanalysisandhematologicaladaptations.
Miceexposedto
Miceexposedto
P
normoxia(n=6) chronichypoxia(n=6)
Arterialbloodgasanalysis
pH
PaO2,mmHg
PaCO2,mmHg
HCO3‐,mmol/L
SaO2,%
Baseexcess,mEq/L
Hematologicaladaptations
Hematocrit,%
Hemoglobin,mmol/L
Erythrocytes,x106
MCV,fL
MCH,pg
MCHC,g/dL
p50(a)
7.28(0.01)
129.8(3.7)
35.1(2.7)
15.9(1.1)
100(0.4)
‐9.7(0.9)
7.11(0.02)
34.3(1.6)
35.0(1.4)
10.7(0.7)
24(2.4)
‐21.2(1.4)
45(1)
9.0(0.3)
10.0(0.2)
45.6(0.3)
0.91 (0.01)
19.8(0.2)
29.7(0.4)
76(1)
14.4(0.1)
13.5(0.2)
56.1(0.8)
1.07(0.02)
19.0(0.3)
51.6(0.9)
<.001
<.001
.96
.002
<.001
<.001
<.001
<.001
<.001
<.001
<.001
.026
<.001
Dataaremeans(SE).Abbreviations:CH,chronichypoxia;MCV,meancorpuscularvolume;MCH,mean
corpuscularhemoglobin,MCHC,meancorpuscularhemoglobinconcentration.
116
CHAPTER6
Recoveryoffoodintakebutlowerbodyweightduringchronichypoxia
Directly after initiation of normobaric hypoxia, a drop in food intake was
observed,whichreacheditsnadirat58%ofstartingfoodintakeonday3(Figure
1A). During days 6 to 10, a complete recovery of food intake was seen, and
thereafter it remained stable throughout the remaining 11 days of the
experiment. Starting body weights were not different between the two groups
(32.6±0.3vs33.1±0.6g,P=.55).Duringdays0to6,bodyweightofmiceexposed
to chronic hypoxia gradually decreased with approximately 13% (Figure 1B).
Even though their food intake reached normal values by day 10, their body
weightsremainedstablylowthroughouttherestoftheexperiment.
Figure 1. Hypoxia reduces food intake and body weight, but only food intake
recovers.
A)Foodintakeaspercentageofstartingfoodintake.Threetofourmicewerehousedpercageanddata
representthemeanfoodintakepermouse,relativetoday0.B)Bodyweightaspercentageofstarting
bodyweight.
Circulatinglevelsofcytokinesandadipokines
Chronic hypoxia increased plasma IL‐6 and PAI‐1 concentrations, whereas it
decreased leptin concentrations. Plasma concentrations of TNF‐α, IL‐1β, CCL5,
MCP‐1,KC,andadiponectinremainedunalteredafterchronichypoxia(Table3).
Loss of adipose tissue and skeletal muscle mass and decreased adipocyte size
afterchronichypoxia
Whereas chronic hypoxia decreased lower limb muscle weights by 10.5±2.5%
(Table 4), the loss of AT mass was much more pronounced (Figure 2A). The
decrease in VAT was significantly greater than that of SCAT (‐65% vs ‐55%,
P<.001).Inline,adipocytesinVATshrunksignificantlymorecomparedtothose
in SCAT after chronic hypoxia (‐52% vs ‐40%, P=.001) (Figure 2B). A strong
correlation was found between AT mass and adipocyte size (VAT: r=0.94,
P<.001;SCAT:r=0.81,P<.001).
117
Table3.Circulatinglevelsofcytokinesandadipokines.
Miceexposedto
Miceexposedto
P
IL‐6,pg/ml
0.46(0.37‐0.87)
1.12(0.72‐2.45)
.044
TNF‐α,pg/ml
25.1(23.5‐61.5)
33.3(25.1‐69.1)
.47
IL‐1β,pg/ml
27.1(15.9‐36.4)
25.7(15.9‐35.7)
.90
KC,pg/ml
10.6(4.1‐13.6)
12.1(5.1‐18.1)
.65
CCL5,pg/ml
139(107‐188)
167(140‐209)
.33
MCP‐1,pg/ml
31.7(25.3‐38.8)
33.5(26.2‐40.5)
.70
Leptin,ng/ml
8.9(6.1‐11.9)
4.6(4.2‐5.1)
.002
Adiponectin,μg/ml
31.8(27.3‐38.7)
32.1(27.4‐43.0)
.90
PAI‐1,ng/ml
1.3(1.1‐1.7)
2.4(2.3‐3.0)
.002
Abbreviations: CCL5, chemokine (C‐C motif) ligand 5; IL, interleukin; KC, keratinocyte‐derived
chemokine; MCP‐1, monocyte chemotactic protein‐1; PAI‐1, plasminogen activator‐inhibitor‐1; TNF‐α,
tumornecrosisfactor‐α.Dataaremediansand95%CIs.
Table4.Peripheralskeletalmuscleweights.
Miceexposedto
Miceexposedto
Gastrocnemiusmuscle,mg
Tibialismuscle,mg
Plantarismuscle,mg
Extensordigitorumlongus
muscle,mg
Soleusmuscle,mg
P
317(2.2)
117(1.3)
41.2(0.8)
27.5(0.4)
276(4.0)
106(1.6)
36.0(0.8)
24.6(0.4)
<.001
<.001
<.001
<.001
21.7(0.4)
20.2(0.6)
.10
Abbreviation:CH,chronichypoxia.Dataaremeans(SE).
Figure 2. Chronic hypoxia reduces adipose tissue weight and adipocyte size in
visceral(VAT)andsubcutaneousadiposetissue(SCAT).
Mice were sacrificed after 21 days of normoxic (n=6) or hypoxic (n=8) conditions and VAT and SCAT
were isolated. A) Bilateral VAT and SCAT weights were significantly reduced during chronic hypoxia,
being more pronounced in VAT. B) Decreased adipocyte size during chronic hypoxia, being more
pronouncedinVAT.*P<.001forwithin‐fatpadcomparisons,#P<.001forbetween‐fatpadcomparisons.
118
CHAPTER6
Chronic hypoxia induces upregulation of hypoxia‐sensitive genes in adipose
tissue
Slc2a1, Vegfa and Pfkfb3 are hypoxia‐inducible genes encoding glucose
transporter‐1 (GLUT‐1), vascular endothelial growth factor‐a (VEGF‐A) and the
glycolysis‐promoting enzyme 6‐phosphofructo‐2‐kinase/fructose‐2,6‐biphosphatase
3 (PFKFB3), respectively [34, 35]. Chronic hypoxia increased expression of
Slc2a1inSCATandVegfaandPfkf3binbothSCATandVAT(Figure3),suggesting
that despite the hematological adaptations described above, oxygen‐delivery to
ATwasinsufficientresultinginareducedATPaO2.Thedegreeofupregulationof
thesegeneswassimilarinVATandSCAT.
Figure 3. Expression of hypoxia‐sensitive genes is increased in adipose tissue
afterchronichypoxiaexposure.
werecollected.mRNAexpressionwascorrectedforastableGeNormfactorandpresentedrelativetothe
levelsinVATofnormoxicmice.ExpressionlevelsofVegfa(encodingVascularendothelialgrowthfactor‐
A, VEGF‐A), Pfkfb3 (encoding 6‐phosphofructo‐2‐kinase/fructose‐2,6‐biphosphatase 3, PFKFB3), and
Scl2a1(encodingglucosetransporter1,GLUT‐1)weremeasured.*P<.05.
ChronichypoxiaattenuatestheinflammatoryprofileinVATandSCAT
AT macrophages are considered to be pivotal players in adipose tissue
inflammation[27].InbothVATandSCATofmiceexposedtochronichypoxia,we
foundsignificantlylowergeneexpressionofkeymarkersrelatedtomacrophage
presence (Emr1) and macrophage recruitment (Ccl2) (Figure 4A). This was
confirmedbylowermacrophagedensityinbothVATandSCAT(Figures4B‐C).
The decreases in macrophage markers and density were not different between
VATandSCAT.CLScountsweretoolowtodrawafirmconclusion,buttendedto
be lower in mice exposed to chronic hypoxia, particularly in VAT (data not
shown).
119
Figure4.Reducedadiposetissuemacrophageinfiltrationafterchronichypoxia.
werecollected.A)DecreasedmRNAlevelsofthemacrophagemarkersEmr1(encodingF4/80),andCcl2
(encoding monocyte chemotactic protein‐1, MCP‐1) were observed in VAT and SCAT after chronic
hypoxia.mRNAexpressionwascorrectedforastableGeNormfactorandpresentedrelativetothelevels
in VAT of normoxic mice. B) Adipose tissue macrophage density was assessed by quantification of
immunohistochemically stained macrophages. A significant decrease was observed in AT of mice
exposedtochronichypoxiaC)TherepresentativepicturesillustratedecreasedATmacrophagedensity
andadipocytesizeafterexposuretochronichypoxia(at200×magnification).*P<.05.
In addition, gene expression of Tnf, Il10 and Lep was lower in VAT of mice
exposed to chronic hypoxia, whereas expression of Adipoq and Serpine2 was
increased (Figure 5A). SCAT of mice exposed to chronic hypoxia was further
characterized by lower expression of Il10 and Lep, and an increased Serpine2
120
CHAPTER6
expression (Figure 4B). The decrease in Lep and the increase in Adipoq
expression were more pronounced in VAT than in SCAT. Additionally, the
decrease in Il10 expression was greater in SCAT compared to VAT. The lower
expression of Il1b in mice exposed to chronic hypoxia did not reach statistical
significance,andnochangeswereobservedinIl6expression.
Chronic hypoxia increases expression of genes involved in oxidative and lipid
metabolisminadiposetissue
Wefoundsignificantlyhigherexpressionofregulatorsandmarkersofoxidative
metabolism in the AT of mice exposed to chronic hypoxia (Figure 5B). VAT of
mice exposed to chronic hypoxia was characterized by increased Ppargc1a,
Ppargc1b, Tfam, Adrb3, Cyc1, Ucp1 and Cidea expression. Congruently, SCAT of
mice exposed to chronic hypoxia was characterized by increased Ppargc1b,
Adrb3, Cyc1 and Ucp1 expression. The expression of the adipogenesis marker
Pparg and of the lipogenesis marker Srebp1c were unchanged, whereas
expressionofthelipolysismarkers LipeandPnpla2wasincreasedinbothVAT
and SCAT of mice exposed to chronic hypoxia (Figure 5C). Furthermore, the
expressionofPck1,encodingphosphoenolpyruvatecarboxykinase(PEPCK),was
significantlyincreasedinbothVATandSCATofthesemice(Figure5C).Noneof
the chronic hypoxia‐induced changes in metabolic gene expression were
differentbetweenVATandSCAT.
The alterations in metabolic gene expression after chronic hypoxia, in
particularthemarkedlyincreasedUcp1expression,suggest‘browning’ofthese
white adipose tissue pads. Interestingly, we found that the adipose tissue pads
wereindeedcharacterizedbyabrownappearance(Figure6).
Discussion
Inmiceexposedtochronichypoxia,wefoundconsistentevidenceofdecreased
AT inflammation and alterations in metabolic AT gene expression, suggesting
increased oxidative metabolism and enhanced lipolysis. In view of the pro‐
inflammatoryandoxidative‐to‐glycolyticshiftsthathavebeendocumentedafter
exposingadipocytestoacuteandseverehypoxiainvitro,andofstudiesinobese
micesuggestingalinkbetweenAThypoxiaandlocalinflammation,ourfindings
may be unexpected. However, the present findings are in line with previous
reports, showing that abdominal subcutaneous AT PaO2 was significantly
increased rather than decreased inobese insulin resistant individuals, and was
positively associated with AT gene expression of several pro‐inflammatory
markers [5]. In that study, the lower AT PaO2 in obesity seemed to be due to
decreasedinvivoabdominalsubcutaneousAToxygenconsumption,suggestinga
lower metabolic rate of AT in obese subjects. Furthermore, abdominal
subcutaneous AT PaO2 was inversely associated with AT expression of
121
Figure 5. Significant alterations in adipose tissue gene expression of
adipocytokines and of key markers of oxidative and lipid metabolism after
chronichypoxia.
werecollected.A)mRNA levels ofspecificcytokines andadipokines. B)mRNAlevelsof regulatorsand
markersofoxidativemetabolisminadiposetissueofmiceexposedtochronichypoxia.C)mRNAlevelsof
lipolysis‐related markers. mRNA expression was corrected for a stable GeNorm factor and presented
relative to the levels in VAT of normoxic mice. *P<.05 for within‐fat pad comparisons, #P<.05 for
between‐fatpadcomparisons.
122
CHAPTER6
Figure 6. Brown appearance of visceral and subcutaneous adipose tissue after
chronichypoxia.
Pictures show representative adipose tissue pads of mice exposed to 21 days of normoxia or hypoxia.
Pictures were taken once the tissues were embedded in paraffin. Note the brown appearance of the
adipose tissue pads of mice exposed to chronic hypoxia. A) Visceral adipose tissue. B) Subcutaneous
adiposetissue.
PPARGC1AandVEGFA[5].Thepresentdatasuggestthattheseassociationsmay
also hold in a model of chronic hypoxia. Congruently, it has recently been
reported that primary human SCAT‐derived adipocytes from overweight
individuals displayed reduced expression of nuclear factor‐κB related genes
(includingCCL2),andshowedabluntedresponsetoTNF‐αstimulationinterms
of CCL2 expression and protein secretion when exposed to hypoxia [20]. Our
data underline the complexity of the effects of hypoxia on AT function and
suggestimportantdifferencesbetweenacuteandchroniceffects.Tothebestof
ourknowledge,nopreviousstudyhasinvestigatedATfunctionafterexposureto
chronichypoxia,precludingdirectcomparisons.
Thegrossphenotypicalchangesinmiceexposedtochronichypoxiashould
not be disregarded when interpreting AT inflammatory and metabolic
alterations. Chronic hypoxia markedly reduced fat mass and, congruently,
decreased adipocyte size. This may have contributed to decreased AT
inflammation. Indeed, adipocyte size has been recognized as an important
determinantofadipokineexpressionandsecretion,withlargeadipocytesbeing
more pro‐inflammatory than small adipocytes [36]. During the course of
hypertrophy, adipocytes produce CCL2, and fatty acids released from
hypertrophied adipocytes can bind Toll‐like receptor 4 complex, thereby
activating an inflammatory response in AT macrophages [37]. Conversely, our
data on smaller adipocytes, decreased AT Ccl2 and Emr1 expression and
decreased AT macrophage density suggest that these mechanisms of AT
macrophageinfiltrationandactivationaresignificantlyattenuatedafterchronic
hypoxia.
Low‐grade systemic inflammation, associated with elevated IL‐6, has been
suggested to play a pivotal role in both obesity and COPD, and AT has been
considered as an important contributor in both diseases [38, 39]. It has
123
previously been shown that sustained hypoxia results in increased levels of
circulating IL‐6 [40, 41], which is confirmed by the current data. Also, chronic
hypoxiaalmostdoubledtheconcentrationsofcirculatingPAI‐1,whichisamain
inhibitoroffibrinolysis.Thisisinlinewithapreviousstudy[42],andwiththe
well‐known association between systemic inflammation and a pro‐thrombotic
state [43]. The changes in circulating PAI‐1 and leptin concentrations after
chronic hypoxia were paralleled by changes in AT gene expression of these
adipokines. However, AT IL‐6 gene expression was unchanged and other pro‐
inflammatorymarkerswereevendecreased.Thissuggeststhatchronichypoxia‐
enhanced circulating IL‐6 levels and likely chronic hypoxia‐induced low‐grade
systemicinflammationingeneral,arenotaccountedforbytheAT.
In our model, exposure to hypoxia initially induced significant weight loss
anddecreasedfoodintake,butaftertendaysfoodintakecompletelynormalized
whereas body weight remained stably low. These data suggest that a new
homeostatic equilibrium was established which is characterized by a
hypermetabolic state. Interestingly, this was accompanied by major changes in
ATmetabolicgeneexpressionthatcollectivelysuggestincreasedmitochondrial
biogenesis,mitochondrialfunctionandlipolysisafterchronichypoxia.Giventhis
specificpatternofmetabolicgeneexpression,itistemptingtospeculatethatAT
remodeling occurred via chronic stimulation of adipocytic β3‐adrenergic
receptors (β3‐AR). Indeed, nor‐adrenergic remodeling of white AT is
characterized by an elevation of the metabolic rate, an expansion of
mitochondrialmassandupregulationoffattyacidoxidationgenes,andhaseven
beenshowntoinduceuncouplingprotein‐1(UCP‐1),whichistypicallyabrown
AT marker [44, 45]. Hormone sensitive lipase‐mediated lipolysis has, notably,
been recognized asan important component of AT remodeling followingβ3‐AR
activation [45]. Additionally, PEPCK (encoded by Pck1) is a key mediator of
glyceroneogenesis, which involves re‐esterification of fatty acids to
triacylglycerol serving adipocytic retention rather than secretion of fatty acids
[46]. In brown AT, it has been suggested that glyceroneogenesis regulates the
deliveryofFAtothemitochondria,allowingforUCP‐1activation[46].Moreover,
Pck1 transcription is induced by nor‐adrenergic stimulation in brown AT [47]
and in white AT pro‐inflammatory cytokines inhibit Pck1 expression [48]. β3‐
adrenergic remodeling is induced by chronic cold exposure [44], therefore, a
limitation of the present study is that we did not assess body temperature nor
applyathermoneutralenvironment.Forthesereasons,wecannotfullyexclude
that part of our findings can be explained by relative cold exposure in mice
exposedtochronichypoxia.AsmetabolicremodelingofwhiteATisreceivinga
great deal of attention in the combat against obesity, our finding of this
phenomenoncoincidingwithdecreasedATinflammationmayberelevanttothe
obesityfield.Thelowerfatmassandthemarkedlyreducedadipocytesizepoint
towardsincreasedlipolysisafterchronichypoxia,whichisfurthersupportedby
the increased gene expression of the key lipolytic enzymes HSL and ATGL, and
124
CHAPTER6
theunchangedgeneexpressionoftheadipogenic/lipogenicmarkersPPAR‐γand
SREBP‐1c.Alimitationofthisstudy,however,isthatwedidnotassesssystemic
markersoflipolysisduetolimitedavailabilityofmaterial.
In conclusion, chronic hypoxia is associated with decreased rather than
increasedATinflammation,andmarkedlydecreasedfatmassandadipocytesize.
Furthermore, our data indicate that chronic hypoxia is accompanied by
alterations in AT metabolic gene expression, pointing towards an enhanced AT
metabolicrate.
125
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
126
Buist AS, McBurnieMA, VollmerWM, GillespieS,Burney P, Mannino DM, Menezes AM, Sullivan SD, Lee TA,
Weiss KB, et al. International variation in the prevalence of COPD (the BOLD Study): a population‐based
prevalencestudy.Lancet.2007;370(9589):741‐50.
Popkin BM. Recent dynamics suggest selected countries catching up to US obesity. Am J Clin Nutr.
2010;91(1):284S‐8S.
Franssen FM, O'Donnell DE, Goossens GH, Blaak EE, Schols AM. Obesity and the lung: 5. Obesity and COPD.
Thorax.2008;63(12):1110‐7.
Vozoris N, O'Donnell DE. Prevalence, risk factors, activity limitation and health care utilization of an obese,
population‐basedsamplewithchronicobstructivepulmonarydisease.CanRespirJ.2012;19(3):e18‐24.
Goossens GH, Bizzarri A, Venteclef N, Essers Y, Cleutjens JP, Konings E, Jocken JW, Cajlakovic M, RibitschV,
ClementK,etal.Increasedadiposetissueoxygentensioninobesecomparedwithleanmenisaccompaniedby
insulinresistance,impairedadiposetissuecapillarization,andinflammation.Circulation.2011;124(1):67‐76.
GoossensGH,BlaakEE.Adiposetissueoxygentension:implicationsforchronicmetabolicandinflammatory
diseases.CurrOpinClinNutrMetabCare.2012;15(6):539‐46.
PasaricaM,SeredaOR,RedmanLM,AlbaradoDC,HymelDT,RoanLE,RoodJC,BurkDH,SmithSR.Reduced
adipose tissue oxygenation in human obesity: evidence for rarefaction, macrophage chemotaxis, and
inflammationwithoutanangiogenicresponse.Diabetes.2009;58(3):718‐25.
TrayhurnP.Hypoxiaandadiposetissuefunctionanddysfunctioninobesity.PhysiolRev.2013;93(1):1‐21.
Raguso CA, Luthy C. Nutritional status in chronic obstructive pulmonary disease: role of hypoxia. Nutrition.
2011;27(2):138‐43.
ScholsAM.Pulmonarycachexia.IntJCardiol.2002;85(1):101‐10.
Wagner PD. Possible mechanisms underlying the development of cachexia in COPD. Eur Respir J.
2008;31(3):492‐501.
GoossensGH.Theroleofadiposetissuedysfunctioninthepathogenesisofobesity‐relatedinsulinresistance.
PhysiolBehav.2008;94(2):206‐18.
2004;92(3):347‐55.
HosogaiN,FukuharaA,OshimaK,MiyataY,TanakaS,SegawaK,FurukawaS,TochinoY,KomuroR,Matsuda
M, et al. Adipose tissue hypoxia in obesity and its impact on adipocytokine dysregulation. Diabetes.
2007;56(4):901‐11.
Rausch ME, Weisberg S, Vardhana P, Tortoriello DV. Obesity in C57BL/6J mice is characterized by adipose
tissuehypoxiaandcytotoxicT‐cellinfiltration.IntJObes(Lond).2008;32(3):451‐63.
YeJ,GaoZ,YinJ,HeQ.Hypoxiaisapotentialriskfactorforchronicinflammationandadiponectinreductionin
adiposetissueofob/obanddietaryobesemice.AmJPhysiolEndocrinolMetab.2007;293(4):E1118‐28.
MazzattiD,LimFL,O'HaraA,WoodIS,TrayhurnP.Amicroarrayanalysisofthehypoxia‐inducedmodulation
ofgeneexpressioninhumanadipocytes.ArchPhysiolBiochem.2012;118(3):112‐20.
WreeA,MayerA,WestphalS,BeilfussA,CanbayA,SchickRR,GerkenG,VaupelP.Adipokineexpressionin
brownandwhiteadipocytesinresponsetohypoxia.JEndocrinolInvest.2012;35(5):522‐7.
Yu J, Shi L, Wang H, Bilan PJ, Yao Z, Samaan MC, He Q, Klip A, Niu W. Conditioned medium from hypoxia‐
treatedadipocytesrendersmusclecellsinsulinresistant.EurJCellBiol.2011;90(12):1000‐15.
FamullaS,HorrighsA,CramerA,SellH,EckelJ.Hypoxiareducestheresponseofhumanadipocytestowards
TNFalpharesultinginreducedNF‐kappaBsignalingandMCP‐1secretion.IntJObes(Lond).2012;36(7):986‐92.
DiGirolamo M, Newby FD, Lovejoy J. Lactate production in adipose tissue: a regulated function with extra‐
adiposeimplications.FASEBJ.1992;6(7):2405‐12.
PerezdeHerediaF,WoodIS,TrayhurnP.Hypoxiastimulateslactatereleaseandmodulatesmonocarboxylate
transporter(MCT1,MCT2,andMCT4)expressioninhumanadipocytes.PflugersArch.2010;459(3):509‐18.
Yin J, Gao Z, He Q, Zhou D, Guo Z, Ye J. Role of hypoxia in obesity‐induced disorders of glucose and lipid
metabolisminadiposetissue.AmJPhysiolEndocrinolMetab.2009;296(2):E333‐42.
Quintero P, Gonzalez‐Muniesa P, Garcia‐Diaz DF, Martinez JA. Effects of hyperoxia exposure on metabolic
markersandgeneexpressionin3T3‐L1adipocytes.JPhysiolBiochem.2012;68(4):663‐9.
Increased adipose tissue expression of proinflammatory CD40, MKK4 and JNK in patients with very severe
chronicobstructivepulmonarydisease.Respiration.2011;81(5):386‐93.
EffectsofHypoxiaonAdiposeTissueExpressionofNFkappaB,IkappaBalpha,IKKgammaandIKAPinPatients
withChronicObstructivePulmonaryDisease.CellBiochemBiophys.2012.
MorrisDL,SingerK,LumengCN.Adiposetissuemacrophages:phenotypicplasticityanddiversityinleanand
obesestates.CurrOpinClinNutrMetabCare.2011;14(4):341‐6.
CHAPTER6
28. Sabo‐Attwood T, Ramos‐Nino M, Bond J, Butnor KJ, Heintz N, Gruber AD, Steele C, Taatjes DJ, Vacek P,
MossmanBT.GeneexpressionprofilesrevealincreasedmClca3(Gob5)expressionandmucinproductionina
murinemodelofasbestos‐inducedfibrogenesis.AmJPathol.2005;167(5):1243‐56.
29. LubinJH,ColtJS,CamannD,DavisS,CerhanJR,SeversonRK,BernsteinL,HartgeP.Epidemiologicevaluation
ofmeasurementdatainthepresenceofdetectionlimits.EnvironHealthPerspect.2004;112(17):1691‐6.
30. WongHL,PfeifferRM,FearsTR,VermeulenR,JiS,RabkinCS.Reproducibilityandcorrelationsofmultiplex
cytokinelevelsinasymptomaticpersons.CancerEpidemiolBiomarkersPrev.2008;17(12):3450‐6.
31. Helsel DR. Fabricating data: how substituting values for nondetects can ruin results, and what can be done
aboutit.Chemosphere.2006;65(11):2434‐9.
32. VandesompeleJ,DePreterK,PattynF,PoppeB,VanRoyN,DePaepeA,SpelemanF.Accuratenormalization
2002;3(7):RESEARCH0034.
33. PoggiM,EngelD,ChristA,BeckersL,WijnandsE,BoonL, DriessenA,CleutjensJ,WeberC,GerdesN,etal.
CD40L deficiency ameliorates adipose tissue inflammation and metabolic manifestations of obesity in mice.
ArteriosclerThrombVascBiol.2011;31(10):2251‐60.
34. ObachM,Navarro‐SabateA,CaroJ,KongX,DuranJ,GomezM,PeralesJC,VenturaF,RosaJL,BartronsR.6‐
Phosphofructo‐2‐kinase(pfkfb3)genepromotercontainshypoxia‐induciblefactor‐1bindingsitesnecessary
fortransactivationinresponsetohypoxia.JBiolChem.2004;279(51):53562‐70.
35. Trayhurn P, WangB, Wood IS. Hypoxia in adipose tissue: a basis for the dysregulation of tissue function in
obesity?BrJNutr.2008;100(2):227‐35.
36. SkurkT,Alberti‐HuberC,HerderC,HaunerH.Relationshipbetweenadipocytesizeandadipokineexpression
andsecretion.JClinEndocrinolMetab.2007;92(3):1023‐33.
37. ItohM,SuganamiT,HachiyaR,OgawaY.Adiposetissueremodelingashomeostaticinflammation.IntJInflam.
2011;2011:720926.
38. EderK,BaffyN,FalusA,FulopAK.Themajorinflammatorymediatorinterleukin‐6andobesity.InflammRes.
2009;58(11):727‐36.
39. vandenBorstB,GoskerHR,KosterA,YuB,KritchevskySB,LiuY,MeibohmB,RiceTB,ShlipakM,YendeS,et
al.Theinfluenceofabdominalvisceralfatoninflammatorypathwaysandmortalityriskinobstructivelung
disease.AmJClinNutr.2012;96(3):516‐26.
40. Kozak W, Wrotek S, Walentynowicz K. Hypoxia‐induced sickness behaviour. J Physiol Pharmacol. 2006;57
Suppl8:35‐50.
41. RajaKB,GOL‐D,PetersTJ,McKieAT,SimpsonRJ.Roleofinterleukin‐6inhypoxicregulationofintestinaliron
absorption.BrJHaematol.2005;131(5):656‐62.
42. PinskyDJ,LiaoH,LawsonCA,YanSF,ChenJ,CarmelietP,LoskutoffDJ,SternDM.Coordinatedinductionof
plasminogenactivatorinhibitor‐1(PAI‐1)andinhibitionofplasminogenactivatorgeneexpressionbyhypoxia
promotespulmonaryvascularfibrindeposition.JClinInvest.1998;102(5):919‐28.
43. EsmonCT.Theinteractionsbetweeninflammationandcoagulation.BrJHaematol.2005;131(4):417‐30.
44. Lowell BB, Spiegelman BM. Towards a molecular understanding of adaptive thermogenesis. Nature.
2000;404(6778):652‐60.
45. MottilloEP,ShenXJ,GrannemanJG.Roleofhormone‐sensitivelipaseinbeta‐adrenergicremodelingofwhite
adiposetissue.AmJPhysiolEndocrinolMetab.2007;293(5):E1188‐97.
46. HansonRW,ReshefL.Glyceroneogenesisrevisited.Biochimie.2003;85(12):1199‐205.
47. HahnP,KirbyLT.Theeffectsofcatecholamines,glucagon,anddietonenzymeactivitiesinbrownfatandliver
oftherat.CanJBiochem.1974;52(9):739‐43.
48. FeingoldKR,MoserA,ShigenagaJK,GrunfeldC.Inflammationinhibitstheexpressionofphosphoenolpyruvate
carboxykinaseinliverandadiposetissue.InnateImmun.2012;18(2):231‐40.
127
CHAPTER7
DietaryfiberandfattyacidsinCOPD
risk and progression: a systematic
review
Eric L.A. Fonseca Wald, Bram van den Borst, Harry R. Gosker,
AnnemieM.W.J.Schols
Submitted
129
DIETARYFIBERANDFATTYACIDSINCOPD
Abstract
Background: Non‐smoking lifestyle factors including dietary components have
attracted increasing attention in the risk for and course of chronic obstructive
pulmonarydisease(COPD).Dietaryfiberandfattyacidshavebeenhypothesized
to play an immunomodulating role in chronic inflammatory diseases, including
COPD.
Objective:Toreviewthecurrentevidenceonthepotentialrolesofdietaryfiber
orfattyacidsinCOPDriskandprogression.
Methods: We systematically searched Pubmed, EMBASE, Cochrane
Collaboration Database and conference databases for original studies in adults
addressingtheassociationbetweenfiberorfattyacidintakeandCOPDinterms
ofrisk,lungfunctionandrespiratorysymptoms.
Results:Ninearticleswereincludedofwhichfourreportedondietaryfiberand
fiveonfattyacids.Dataofstudiescouldnotbepooledbecauseofmethodological
diversity. Greater intake of dietary fiber has been consistently associated with
reduced COPD risk, better lung function, and reduced respiratory symptoms.
ResultsontheassociationsbetweenfattyacidsandCOPDareinconsistent.
Conclusions: Dietary quality, particularly in terms of dietary fiber amount,
deservesfurtherattentioninCOPDpreventionandmanagementprograms.
130
CHAPTER7
Introduction
An estimated 64 million people suffer from chronic obstructive pulmonary
disease(COPD)worldwide,adiseasethatisassociatedwithhighmorbidityand
mortality [1]. COPD has been defined as a preventable and largely lifestyle‐
induced disease characterized by an abnormal pulmonary inflammatory
response resulting in a progressive, partially irreversible airflow limitation [2].
In addition, COPD has been shown to have a profound systemic component
characterizedbylow‐gradesystemicinflammationandco‐morbidities[2].
ThediseasestateofCOPDislikelytheproductofgeneticandenvironmental
factors [3]. Although exposure to cigarette smoke is considered the principal
environmental risk factor, an estimated 25‐45% of COPD patients have never
smoked[4],only~20%ofsmokerseventuallydevelopCOPD[5],andonly~10%
of the variability in forced expiratory volume in 1s (FEV1) is explained by
smokinginthegeneralpopulation[6].Ithasthereforebeensuggestedthatnon‐
smokingenvironmentalfactorssuchasdietandphysicalactivityshouldalsobe
considered.Foracomprehensiveoverviewontheeffectsofphysicalinactivityon
COPD risk we refer the reader to a recent review by Hopkinson et al. [7]. The
presentsystematicreviewwasundertakentosummarizethecurrentknowledge
ontheeffectsofspecificdietarycomponents,morespecificallydietaryfiberand
fattyacids,onCOPDriskandprogression.
NutritionalsupportinCOPDmanagementcurrentlyhasaprimaryfocuson
malnourished patients. In recent years, however, nutritional research in COPD
has also explored associations between dietary patterns and specific dietary
components with COPD risk and progression. It was shown that a diet rich in
fruit,vegetables,whole‐mealcerealsandfishreducestheriskofCOPDwhereasa
“Westerndiet”richinrefinedgrains,curedandredmeats,dessertsandFrench
frieshasbeenshowntoincreaseCOPDrisk[8‐10].However,knowledgeonthe
influence of specific dietary components may ultimately provide the best
possible way to optimize dietary recommendations in COPD. Dietary fiber and
fattyacidshavebeenshowntoinfluencetheimmunesysteminvariouswaysand
couldbehypothesizedtoplayaroleinchronicinflammatorydiseasesincluding
COPD. Greater intake of dietary fiber has been associated with improved gut
immunity and reduced systemic inflammation [11‐14]. Polyunsaturated fatty
acids (PUFAs) have been shown to possess pro‐ and anti‐inflammatory
properties through direct influences on membrane functioning and through
binding with peroxisome proliferator‐activated receptors (PPARs), and
indirectlyviatheirmetabolizationintoeicosanoids[15].
The present systematic review summarizes the current evidence on the
potential roles of dietary fiber or specific fatty acids in COPD risk and
progression.
131
Methods
Datasourcesandsearchstrategy
This systematic review was performed according to MOOSE (Meta‐analysis Of
Observational Studies in Epidemiology) guidelines [16]. Relevant articles were
searched in Pubmed (through August 2012), EMBASE (1989 through August
2012), Cochrane Collaboration Database (through August 2012), and in
conference databases of the European Respiratory Society (2006‐2011),
AmericanThoracicSociety(2008‐2012),EuropeanSocietyforClinicalNutrition
and Metabolism (2003‐2012), American Society for Parenteral and Enteral
Nutrition (2009‐2011) and Experimental Biology (2010‐2012). Databases were
either systematically searched by combining free terms and subject headings
(MeshorEmtreetermsforPubmedandEMBASE,respectively)whenavailable,
orwerehandsearched.Thesearchstrategyconsistedoftermsondietaryfiber
and specific fatty acids in combination with relevant COPD‐related terms and
lung function (see Appendix for terms used and search strategies per e‐
database). In Pubmed and EMBASE the limits ‘humans’ and ‘adults’ were
imposed. Additionally, references of retrieved articles were hand searched to
findmorearticles.
Studyselection
Articles identified by the search were first screened by title. Subsequently, the
corresponding abstracts of potentially relevant hits were independently
screened by two researchers (ELAFW and BB) based on selection criteria. The
second researcher was blinded for journal, authors, title, publication date and
publication language (all abstracts were in English). In case of disagreement,
consensuswasreachedontheselectionofstudiesforfull‐textassessment.Tobe
included,studieshadtomeetthefollowingcriteria:studiesneededtobeoriginal
studiesinadultsaddressingtheassociationbetweendietaryfiberorfattyacids
withpredefinedoutcomemeasures(pulmonaryfunction,pulmonarysymptoms
and/or COPD risk/ incidence/ prevalence). Studies that assessed plasma
biomarkers of specific fatty acids as surrogate marker of dietary intake were
excluded. In case of uncertainty for inclusion of full‐text articles or conference
abstracts, consensus was reached via a second reviewer. Native speakers were
consultedforforeignlanguagearticles.
132
CHAPTER7
Results
Searchresults
A flowchart of study selection is presented in Figure 1. Briefly, the search in
electronic databases yielded 5016 hits from Pubmed, 5088 from EMBASE and
218fromtheCochraneCollaborationDatabase,fromwhichintotal220possibly
eligiblearticleswerefound.Additionally,16articleswereidentifiedinreference
lists. In total, 236 abstracts were judged on the need to assess the full‐text.
Subsequently, 88 full‐text articles and 41 conference abstracts underwent
thoroughassessmentbasedonthepredefinedcriteria.Finally,ninearticleswere
included.Ofthese,fourreporteddataonassociationsbetween fiberintakeand
COPD,andfiveinvestigatedassociationsbetweenfattyacidintakeandCOPD.At
first, the objective was to perform a meta‐analysis. However, the studies
provedinsufficientlyhomogeneoustoallowpoolingofresultsortabulatingtheir
respective data. Hence, we provide a narrative synthesis of the respective
studies.
Allthestudiesusedafoodfrequencyquestionnaireorvalidatedanaloguesto
assess nutrient intake. In addition, all the studies adjusted extensively for
possibleconfoundingbyknownriskfactors,includingsmoking.
Figure1.Flowchartofarticleselection.
133
AssociationsbetweendietaryfiberandCOPD
Inthelargepopulation‐basedAtherosclerosisRiskinCommunities(ARIC)Study
comprising 11.897 US men and women aged 44‐66y, higher total dietary fiber
intakewascross‐sectionallyassociatedwithbetterlungfunction(FEV1,FVCand
FEV1/FVCratio)[17].Notably,peopleinthehighestquintileoftotaldietaryfiber
intake (median 25.0 g/day) had an adjusted 60 ml higher FEV1 compared with
people in the lowest quintile (median 10.2 g/day). In addition, higher total
dietary fiber intake was associated with lower odds ratios for COPD (OR 0.85,
95%CI0.68‐1.05).COPDwasdefinedaspre‐bronchodilatorFEV1/FVC<0.7and
FEV1<80%ofpredictedand/orself‐reportedpersistedcoughandproductionof
phlegmonmostdaysforatleastthreeconsecutivemonthsoftheyearfortwoor
moreyears.Similardatawereobtainedwheninvestigatedforfiberfromcereals
andfiberfromfruitbutnotforfiberfromvegetables.
Another study by Hirayama et al. [18] compared n=278 Japanese COPD
patients(meanFEV157%ofpredicted)andn=340community‐basednon‐COPD
controls and found that the mean vegetable and fruit intakes of COPD patients
were significantly lower. The authors subsequently applied logistic regression
analyses to estimate the odds of having COPD across quartiles of nutrient
intakes. They found that people who consumed ≥16.1 g total dietary fiber per
day had lower odds of having COPD compared to those who consumed <8.8 g
totaldietaryfiberperday(OR0.49,95%CI0.26‐0.94).Uponstratificationinto
soluble and insoluble fiber intake, the authors found a similar association for
insolublebutnotsolublefiber.
Although the above referenced data from the ARIC study and the Japanese
studymaysuggestarolefordietaryfiberinCOPDetiology,theircross‐sectional
natures preclude from drawing any causal inferences. Two other studies,
however, applied a prospective design and showed remarkably consistent
results.First,ina5‐yearprospectivepopulation‐basedcohortof63.257middle‐
agedmenandwomenresidinginChinaandSingapore[19],higherintakeofnon‐
starch polysaccharides was negatively associated with the incidence of “COPD
symptoms” (defined as cough and phlegm, both shorter and longer than 3
months).Interestingly,whereasfruitandvitaminCintakeswerealsoassociated
with incident COPD symptoms, these associations disappeared after further
adjustmentfordietaryfiberintake.Thissuggeststhatdietaryfiberintakemaybe
atleastpartlyresponsiblefortheseassociations.Itshouldbetakenintoaccount
that no pulmonary function data were available in this study, and that the
respiratory symptoms designated as “COPD symptoms” may have reflected
respiratory conditions other than COPD. Another 16‐year prospective study by
Varrasoetal.[20]in111.580USmenandwomenfromtheNurses’HealthStudy
and Health Professionals Follow‐up Study identified incident COPD cases by
means of a thorough questionnaire and required report of a diagnostic test at
diagnosis. Furthermore, the epidemiologic COPD definition was validated in a
134
CHAPTER7
randomsampleandshowedconfirmationofdiagnosisinupto88%.Itwasfound
that compared to people in the lowest quintile of total dietary fiber intake
(median 11.2 g/day) those in the highest quintile (median 28.4 g/day) had a
lower adjusted relative risk of newly diagnosed COPD (RR 0.67, 95% CI 0.50‐
0.90).Uponstratificationondietaryfibertype,onlycereal(primarilyinsoluble)
fiberintakewasindependentlyassociatedwithnewlydiagnosedCOPD.
AssociationsbetweenfattyacidsandCOPD
Shaharetal.[21]investigatedthecross‐sectionalassociationsbetweenn‐3fatty
acidintake(eicosapentaenoicacid[EPA]anddocosahexaenoicacid[DHA])and
lungfunctionandCOPDstatusinsubjectsparticipatinginthepopulation‐based
ARIC Study. It was found that higher fish consumption but also specifically
higher total n‐3 fatty acid intake was strongly associated with lower odds for
COPDandlowerlungfunction(dose‐dependently).Nosignificantfindingswere
obtained in analyses restricted to never‐smokers in this study. N‐6 fatty acid
intakewasnotreportedinthisstudy.
In the cross‐sectional population‐based MORGEN‐EPIC study [22], data on
intake of specific fatty acids and lung function were available of 13.820 Dutch
men and women aged 20‐59y. The prevalence of COPD (GOLD 2 or higher, no
post‐bronchodilator data) was 4%. Of the n‐3 fatty acids, only DHA was
significantly associated with FEV1, but the direction of this association was
inverse. Also, higher intake of n‐6 fatty acids were generally associated with
lower FEV1 and greater ORs for COPD. Moreover, it was found that smoking
statussignificantlyinteractedwiththeassociationsbetweenn‐6fattyacidsand
FEV1; the inverse associations were stronger in smokers compared to non‐
smokers. In the entire population, α‐linolenic acid, EPA, docosapentaenoic acid
and DHA (all n‐3) intake were all associated with an increased risk of wheeze,
andEPA(n‐3),eicosadienoicacid(n‐6)andeicosatrienoicacid(n‐6)intakewere
positivelyassociatedwithrespiratorysymptoms(chroniccough,chronicphlegm
and/orbreathlessness).Finally,COPDprevalencewasnegativelyassociatedwith
docosapentainoic acid (n‐3) and docosapentaenoic acid (n‐6), and positively
with eicosadienoic acid (n‐6), eicosatrienoic acid (n‐6), arachidonic acid (n‐6)
anddocosatetraenoicacid(n‐6)intake.
In the much smaller population‐based Hunter Community Study [23],
comprising n=195 Australian men and women aged 55‐85y (including n=45
subjects with FEV1<80% of predicted), no cross‐sectional associations were
found between dietary intake of saturated fatty acids, monounsaturated fatty
acids,n‐3polyunsaturatedfattyacids(PUFA)andn‐6‐PUFAswithFEV1,FVCor
FEV1/FVC.
In the same population in which Hirayama et al. reported data on
associationsbetweendietaryfiberandCOPD[18],theauthorsalsoinvestigated
associations for specific fatty acids which were published in a separate paper
135
[24].Comparedtocontrols,COPDpatientsconsumedsignificantlylessPUFAs,n‐
6fattyacidsandn‐3fattyacidsbutsimilarquantitiesofsaturatedfattyacidsand
mono‐unsaturated fatty acids.Significant P valuesfortrends were reported for
COPD ORs and for breathlessness across quartiles of PUFA and n‐6 fatty acid
intakes,butnotofn‐3fattyacidsor(n‐3):(n‐6)ratio.Incontrastwiththeother
papers referenced above, this study excluded fatty acids from cooking oils,
margarine, butter, mayonnaise and salad dressings, and may therefore have
underestimatedactualintakelevels.
Finally,anothercross‐sectionalstudyamong20.322peopleaged>30yfrom
thepopulation‐basedSecondNationalHealthandNutritionExaminationSurvey
(NHANESII)didnotfindanysignificantassociationbetweenoleicacid(n‐9)or
linoleicacid(n‐6)intakewithwheezing[25].
Discussion
There is growing attention for the influences of an overall poor lifestyle in the
risk for and course of COPD, which goes beyond smoking. This review was
undertakentosummarizethecurrentdataontherelationsbetweendietaryfiber
or fatty acids and COPD risk and progression. At present, the available data
suggest a particular important role for dietary fiber in COPD. Indeed, greater
intake of dietary fiber has been associated with better lung function in the
general population, lower oddsof COPD, and lower risk of incident respiratory
symptomsandCOPD.Onthecontrary,thestudiesexaminingassociationsoffatty
acids are limited by their cross‐sectional natures and we must interpret their
dataasbeinginconsistent.Concerningboththeincludedstudiesondietaryfiber
and fatty acids, most of them retrospectively analyzed data of pre‐existing
epidemiological cohorts. Consequently, the methodological differences
precludedperformingameta‐analysis.
Studies have consistently shown that COPD patients consume less dietary
fiber,vegetablesandfruitcomparedwithcontrolsubjects[18,26].Accordingto
the European Food Safety Authority (EFSA) the current recommendation for
dietary fiber intake is 25 g/day, and the actual intake averages between 15‐30
g/dayacrossEuropeancountries[27].Ofthestudieswereviewed,onlythosein
thehighestquartiles(orquintiles)oftwostudiesmetthisrecommendation.The
EFSA also states that dietary fiber intake of >25 g/day may be even more
beneficialasithasbeenconsistentlyshowntoreducetheriskoftype2diabetes,
cardiovascular disease and colorectal cancer [27]. There is currently no data
available whether such amounts of dietary fiber intake are also beneficial in
termsofreducedCOPDrisk,butitmayimprovesystemicmanifestationsofthe
disease.Inarecent3‐yearprospectivestudyin120COPDpatients,participants
were randomized to follow a fruit and vegetable rich diet or their regular diet.
The former group succeeded to increase fruit and vegetable intake and,
interestingly,thiswasassociatedwithaslightimprovementinFEV1over3years
136
CHAPTER7
(by approximately 5% of predicted), whereas the control group was
characterizedbyadeclineinFEV1(byapproximately9%ofpredicted,between‐
groupdifferenceP=.03)[28].Thatstudywasperformedagainstthebackground
ofapotentialbenefitofincreaseddietaryantioxidants;however,theresultsmay
alsohavebeenpartlyaccountedforbythesimultaneousincreaseindietaryfiber
intake. The feasibility of increasing fruitand vegetable intake in COPD patients
wasconfirmedinarecentexploratorystudy[29].
TheARICstudyreporteda60mldifferenceinFEV1betweenpersonsinthe
lowest vs highest quintiles of fiber intake, in favor of the latter [17]. This is a
clinically substantial difference since normal FEV1 decline in adults is
approximately 30 ml/y [30], and the FEV1 difference between never smokers
and current smokers is generally around 200 ml [31]. Dietary fiber may
thereforebeanimportantmodifiablefactortopreventordelaysignificantlung
functiondecline.
Currentevidenceisnotconclusiveaboutwhichfooditemscontainingdietary
fiber are more inversely associated with COPD. Two studies found significant
associations for fruit and cereal fiber but not for vegetable fiber [17, 20].
However,itisnotclearwhetherthisisaccountabletosolubleorinsolublefiber.
Generally,cerealscontainmainlywater‐insolublefiber,butarealsoarelatively
good source of water‐soluble fiber (about 25% [11]); fruits contain mainly
water‐soluble fiber and vegetables water‐insoluble fiber [32]. On the contrary,
onlyinsolublefiberwasassociatedwithlowerCOPDORsintheJapanesestudy
[18],butthepoorrangeofsolublefiberintakeinthatstudy(≥3.4g/dayvs≤1.6
g/day)isalimitationinthisrespect.
n‐6 fatty acids are thought to be more pro‐inflammatory than their n‐3
counterparts[15],anditishypothesizedthatgreatern‐3fattyacidintakeand/or
lowern‐6:n‐3fattyacidratiosareassociatedwithbeneficialoutcomesrelatedto
COPD. In a cross‐sectional analysis of 250 COPD patients, higher intake of α‐
linolenicacid(n‐3)wasassociatedwithlowercirculatingtumornecrosisfactor‐
α concentrations and higher intake of arachidonic acid (n‐6) was related to
higher interleukin‐6 and C‐reactive protein concentrations [33]. However,
whereas specific dietary fatty acids may be associated with systemic
inflammation in COPD patients, the reviewed articles have produced
inconclusive data with respect to outcome measures related to COPD. In most
studiesn‐3fattyacidintakedoesnotseemtobeassociatedwithCOPD.However,
one cross‐sectional study did show a significant association between n‐3 fatty
acid intake and lung function and COPD risk [21]. The median intake of the
highestquartileofEPAandDHAinthisstudywasconsiderablyhighercompared
tothatintheotherstudies.Nonetheless,inallstudies,n‐3fattyacidintakewas
farbelowthecurrentrecommendationsof1.6g/dayformenand1.1g/dayfor
woman [34]. Protective effects may only become apparent with higher intake
levels,whichmayexplainthelackofsignificantresults.Furthermore,ithasbeen
suggested that a n‐6:n‐3 fatty acid ratio between 1:1 to 4:1 may be most
137
beneficialwhereascurrentratiosareapproximately16:1inWesterndiets[35].
Unfortunately, this hypothesis has not been investigated appropriately with
respect to COPD outcomes. Two studies indicated that smoking status is
importanttoconsiderwheninterpretingassociationsbetweenfattyacidintake
and COPD related outcomes, as the negative associations with n‐6 fatty acids
were stronger in smokers than in non‐smokers [22], and positive associations
with EPA+DHA were found in ever‐smokers [21]. These data may suggest that
lower n‐6:n‐3 ratios may be beneficial particularly in smokers, however, this
hypothesishasnotbeenaddressedinthecurrentliterature.
In conclusion, current data indicate that changes in dietary composition, in
particular with respect to dietary fiber amount, may have beneficial effects in
termsofCOPDriskandprogression.Thereforeitisworthwhiletofurtherstudy
the underlying mechanisms in order to optimize specific nutritional
recommendationsinCOPD.
138
CHAPTER7
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
World Health Organization. The global burden of disease: 2004 update. 2008; Available from
http://www.who.int/healthinfo/global_burden_disease/GBD_report_2004update_full.pdf.
Global Initiative for Chronic Obstructive Lung Disease. Global strategy for the diagnosis, management, and
prevention of chronic obstructive pulmonary disease. Revised 2011. Available from
http://www.goldcopd.org/uploads/users/files/GOLD_Report_2011_Feb21.pdf.
WanES,SilvermanEK.GeneticsofCOPDandemphysema.Chest.2009;136(3):859‐66.
SalviSS,BarnesPJ.Chronicobstructivepulmonarydiseaseinnon‐smokers.Lancet.2009;374(9691):733‐43.
MadisonJM,IrwinRS.Chronicobstructivepulmonarydisease.Lancet.1998;352(9126):467‐73.
Peat JK, Woolcock AJ, Cullen K. Decline of lung function and development of chronic airflow limitation: a
longitudinalstudyofnon‐smokersandsmokersinBusselton,WesternAustralia.Thorax.1990;45(1):32‐7.
Hopkinson NS, Polkey MI. Does physical inactivity cause chronic obstructive pulmonary disease? Clin Sci
(Lond).2010;118(9):565‐72.
ShaheenSO,JamesonKA,SyddallHE,AihieSayerA,DennisonEM,CooperC,RobinsonSM.Therelationshipof
dietary patterns with adult lung function and COPD. Eur Respir J. [Research Support, Non‐U.S. Gov't].
2010;36(2):277‐84.
Varraso R, Fung TT, Barr RG, Hu FB, Willett W, Camargo CA, Jr. Prospective study of dietary patterns and
chronicobstructivepulmonarydiseaseamongUSwomen.AmJClinNutr.2007;86(2):488‐95.
Varraso R, Fung TT, Hu FB, Willett W, Camargo CA. Prospective study of dietary patterns and chronic
GalisteoM,DuarteJ,ZarzueloA.Effectsofdietaryfibersondisturbancesclusteredinthemetabolicsyndrome.
JNutrBiochem.2008;19(2):71‐84.
Ghanim H, Abuaysheh S, Sia CL, Korzeniewski K, Chaudhuri A, Fernandez‐Real JM, Dandona P. Increase in
plasma endotoxin concentrations and the expression of Toll‐like receptors and suppressor of cytokine
signaling‐3inmononuclearcellsafterahigh‐fat,high‐carbohydratemeal:implicationsforinsulinresistance.
DiabetesCare.2009;32(12):2281‐7.
KaczmarczykMM,MillerMJ,FreundGG.Thehealthbenefitsofdietaryfiber:Beyondtheusualsuspectsoftype
2 diabetes mellitus, cardiovascular disease and colon cancer. Metabolism: clinical and experimental.
2012;61(8):1058‐66.
VinoloMA,RodriguesHG,NachbarRT,CuriR.Regulationofinflammationbyshortchainfattyacids.Nutrients.
[ResearchSupport,Non‐U.S.Gov'tReview].2011;3(10):858‐76.
WallR,RossRP,FitzgeraldGF,StantonC.Fattyacidsfromfish:theanti‐inflammatorypotentialoflong‐chain
omega‐3fattyacids.NutrRev.2010;68(5):280‐9.
StroupDF,BerlinJA,MortonSC,OlkinI,WilliamsonGD,RennieD,MoherD,BeckerBJ,SipeTA,ThackerSB.
Meta‐analysis of observational studies in epidemiology: a proposal for reporting. Meta‐analysis Of
ObservationalStudiesinEpidemiology(MOOSE)group.JAMA.2000;283(15):2008‐12.
KanH,StevensJ,HeissG,RoseKM,LondonSJ.Dietaryfiber,lungfunction,andchronicobstructivepulmonary
diseaseintheatherosclerosisriskincommunitiesstudy.AmJEpidemiol.2008;167(5):570‐8.
HirayamaF,LeeAH,BinnsCW,ZhaoY,HiramatsuT,TanikawaY,NishimuraK,TaniguchiH.Dovegetablesand
fruits reduce the risk of chronic obstructive pulmonary disease? A case‐control study in Japan. Prev Med.
2009;49(2‐3):184‐9.
ButlerLM,KohWP,LeeHP,YuMC,LondonSJ.Dietaryfiberandreducedcoughwithphlegm:acohortstudyin
Singapore.AmJRespirCritCareMed.2004;170(3):279‐87.
Varraso R, Willett WC, Camargo CA, Jr. Prospective study of dietary fiber and risk of chronic obstructive
pulmonarydiseaseamongUSwomenandmen.AmJEpidemiol.2010;171(7):776‐84.
ShaharE,FolsomAR,MelnickSL,TockmanMS,ComstockGW,GennaroV,HigginsMW,SorliePD,KoWJ,Szklo
M. Dietary n‐3 polyunsaturated fatty acids and smoking‐related chronic obstructive pulmonary disease.
AtherosclerosisRiskinCommunitiesStudyInvestigators.NEnglJMed.1994;331(4):228‐33.
McKeever TM, Lewis SA, Cassano PA, Ocke M, Burney P, Britton J, Smit HA. The relation between dietary
intakeofindividualfattyacids,FEV1andrespiratorydiseaseinDutchadults.Thorax.2008;63(3):208‐14.
Wood LG, Attia J, McElduff P, McEvoy M, Gibson PG. Assessment of dietary fat intake and innate immune
activationasriskfactorsforimpairedlungfunction.EurJClinNutr.2010;64(8):818‐25.
Hirayama F, Lee AH, Binns CW, Hiramatsu N, Mori M, Nishimura K. Dietary intake of isoflavones and
polyunsaturated fatty acids associated with lung function, breathlessness and the prevalence of chronic
obstructive pulmonary disease: Possible protective effect of traditional Japanese diet. Mol Nutr Food Res.
2010;54(7):909‐17.
SchwartzJ,WeissST.Dietaryfactorsandtheirrelationtorespiratorysymptoms.TheSecondNationalHealth
andNutritionExaminationSurvey.AmJEpidemiol.1990;132(1):67‐76.
vandenBorstB,GoskerHR,KosterA,YuB,KritchevskySB,LiuY,MeibohmB,RiceTB,ShlipakM,YendeS,et
EFSA. Panel on Dietetic Products, Nutrition & Allergies. Scientific Opinion on Dietary Reference Values for
carbohydratesanddietaryfibre.EFSAJ.2010;8:1462‐539.
Keranis E, Makris D, Rodopoulou P, Martinou H, Papamakarios G, Daniil Z, Zintzaras E, Gourgoulianis KI.
Impactofdietaryshifttohigher‐antioxidantfoodsinCOPD:arandomisedtrial.EurRespirJ.2010;36(4):774‐80.
139
29. Baldrick FR, Elborn JS, Woodside JV, Treacy K, Bradley JM, Patterson CC, Schock BC, Ennis M, Young IS,
McKinleyMC.EffectoffruitandvegetableintakeonoxidativestressandinflammationinCOPD:arandomised
controlled trial. The European respiratory journal : official journal of the European Society for Clinical
RespiratoryPhysiology.[ResearchSupport,Non‐U.S.Gov't].2012;39(6):1377‐84.
30. Lee PN, Fry JS. Systematic review of the evidence relating FEV1 decline to giving up smoking. BMC Med.
2010;8:84.
31. OmoriH,NonamiY,MorimotoY.EffectofsmokingonFEVdeclineinacross‐sectionalandlongitudinalstudy
ofalargecohortofJapanesemales.Respirology.2005;10(4):464‐9.
32. Theuwissen E, Mensink RP. Water‐soluble dietary fibers and cardiovascular disease. Physiol Behav.
2008;94(2):285‐92.
33. deBatlleJ,SauledaJ,BalcellsE,GomezFP,MendezM,RodriguezE,BarreiroE,FerrerJJ,RomieuI,GeaJ,etal.
AssociationbetweenOmega3andOmega6fattyacidintakesandseruminflammatorymarkersinCOPD.The
Journalofnutritionalbiochemistry.[ResearchSupport,Non‐U.S.Gov't].2012;23(7):817‐21.
34. FoodandNutritionBoardIoM.Dietaryreferenceintakesforcarbohydrate,fiber,fat,fattyacids,cholesterol,
protein,andaminoacid.Washington,DC:NationalAcademiesPress.2005.
35. SimopoulosAP.Theimportanceoftheomega‐6/omega‐3fattyacidratioincardiovasculardiseaseandother
chronicdiseases.ExpBiolMed(Maywood).2008;233(6):674‐88.
140
CHAPTER7
Appendix:searchtermsusedinonlinedatabases
SearchinPubmedonAugust4,2012yielding5016hits:
("dietary fiber"[All Fields] OR "soluble fiber"[All Fields] OR "insoluble fiber"[All Fields] OR fibre[All
Fields] OR "non starch polysaccharides"[All Fields] OR ("glucans"[MeSH Terms] OR "glucans"[All
Fields]) OR ("inulin"[MeSH Terms] OR "inulin"[All Fields]) OR ("dextrins"[MeSH Terms] OR
"dextrins"[AllFields])OR("lignin"[MeSHTerms]OR"lignin"[AllFields])OR("waxes"[MeSHTerms]OR
"waxes"[AllFields])OR("chitin"[MeSHTerms]OR"chitin"[AllFields])OR("pectins"[MeSHTerms]OR
"pectins"[All Fields]) OR ("oligosaccharides"[MeSH Terms] OR "oligosaccharides"[All Fields]) OR
("gingiva"[MeSH Terms] OR "gingiva"[All Fields] OR "gums"[All Fields]) OR "resistant starches"[All
Fields]OR("dietaryfiber"[MeSHTerms]OR("dietary"[AllFields]AND"fiber"[AllFields])OR"dietary
fiber"[AllFields]OR"roughage"[AllFields])OR("dietaryfiber"[MeSHTerms]OR("dietary"[AllFields]
AND"fiber"[AllFields])OR"dietaryfiber"[AllFields]OR"roughages"[AllFields])ORarabinoxylans[All
Fields]OR("cellulose"[MeSHTerms]OR"cellulose"[AllFields])OR"fattyacids"[AllFields]OR"SFA"[All
Fields] OR "saturated fatty acids"[All Fields] OR "MUFA"[All Fields] OR "monounsaturated fatty
acids"[All Fields] OR "PUFA"[All Fields] OR "polyunsaturated fatty acids"[AllFields]OR"n‐3 fatty"[All
Fields] OR omega‐3[All Fields] OR omega‐6[All Fields] OR "EPA"[All Fields] OR "eicosapentaenoic
acid"[All Fields] OR "DHA"[All Fields] OR "docosahexaenoic acid"[All Fields] OR "ALA"[All Fields] OR
"alpha‐linolenicacid"[AllFields]OR"n‐6fatty"[AllFields]OR"linoleicacid"[AllFields]OR"arachidonic
acid"[All Fields] OR "gamma‐linolenic acid"[All Fields] OR "trans fatty acids"[All Fields] OR
"Hexadecatrienoicacid"[AllFields]OR"Stearidonicacid"[AllFields]OR"Eicosatrienoicacid"[AllFields]
OR "Eicosatetraenoic acid"[All Fields] OR "Clupanodonic acid"[All Fields] OR "Tetracosapentaenoic
acid"[All Fields] OR "Tetracosahexaenoic acid"[All Fields] OR "Eicosadienoic acid"[All Fields] OR
"Dihomogammalinolenicacid"[AllFields]OR(Docosadienoic[AllFields]AND("acids"[MeSHTerms]OR
"acids"[AllFields]OR"acid"[AllFields]))OR"Adrenicacid"[AllFields]OR"Docosapentaenoicacid"[All
Fields]OR"Calendicacid"[AllFields]OR"Myristoleicacid"[AllFields]OR"Palmitoleicacid"[AllFields]
OR (Sapienic[All Fields] AND ("acids"[MeSH Terms] OR "acids"[All Fields] OR "acid"[All Fields])) OR
"Oleic acid"[All Fields] OR "erucic acid"[All Fields] OR "lauric acid"[All Fields] OR "myristic acid"[All
Fields] OR "palmitic acid"[All Fields] OR "stearic acid"[All Fields] OR "arachidic acid"[All Fields] OR
"behenic acid"[All Fields] OR "lignoceric acid"[All Fields] OR "cerotic acid"[All Fields] OR "lard"[All
Fields] OR "butter"[All Fields] OR "coconut oil"[All Fields] OR "palm oil"[All Fields] OR "cottonseed
oil"[All Fields] OR "wheat germ oil"[All Fields] OR "soya oil"[All Fields] OR "olive oil"[All Fields] OR
"cornoil"[AllFields]OR"sunfloweroil"[AllFields]OR"hempoil"[AllFields]OR"canolaoil"[AllFields]
OR "short chain fatty acids"[All Fields] OR "medium chain fatty acids"[All Fields] OR "long chain fatty
acids"[All Fields] OR "rapeseed oil"[All Fields] OR "Dietary Fiber"[Mesh] OR "Polysaccharides"[Mesh]
OR "Dietary fats"[Mesh] OR "Fatty acids"[Mesh] OR "Fish Oils"[Mesh] OR ("fruit"[MeSH Terms] OR
"fruit"[AllFields]OR"fruits"[AllFields])OR("fishes"[MeSHTerms]OR"fishes"[AllFields]OR"fish"[All
Fields]) OR ("vegetables"[MeSH Terms] OR "vegetables"[All Fields]) OR ("meat"[MeSH Terms] OR
"meat"[All Fields]) OR fowl[All Fields] OR ("cereals"[MeSH Terms] OR "cereals"[All Fields] OR
"grain"[All Fields]) OR ("eggs"[MeSH Terms] OR "eggs"[All Fields]) OR ("nuts"[MeSH Terms] OR
"nuts"[AllFields]))AND("lungfunction"[AllFields]OR"pulmonaryfunction"[AllFields]OR"respiratory
function"[All Fields] OR "respiratory symptoms"[All Fields] OR "pulmonary symptoms"[All Fields] OR
("dyspnea"[MeSHTerms]OR"dyspnea"[AllFields]OR"breathlessness"[AllFields])OR("cough"[MeSH
Terms]OR "cough"[All Fields])ORphlegm[AllFields]OR("dyspnoea"[All Fields]OR"dyspnea"[MeSH
Terms] OR "dyspnea"[All Fields]) OR ("dyspnoea"[All Fields] OR "dyspnea"[MeSH Terms] OR
"dyspnea"[All Fields]) OR ("sputum"[MeSH Terms] OR "sputum"[All Fields]) OR FEV[All Fields] OR
FEV1[All Fields] OR "forced expiratory volume"[All Fields] OR "forced vital capacity"[All Fields] OR
FVC[AllFields]OR("pulmonarydisease,chronicobstructive"[MeSHTerms]OR("pulmonary"[AllFields]
AND "disease"[All Fields] AND "chronic"[All Fields] AND "obstructive"[All Fields]) OR "chronic
obstructive pulmonary disease"[All Fields] OR "copd"[All Fields]) OR "chronic obstructive pulmonary
disease"[All Fields] OR ("bronchitis, chronic"[MeSH Terms] OR ("bronchitis"[All Fields] AND
"chronic"[AllFields])OR"chronicbronchitis"[AllFields]OR("chronic"[AllFields]AND"bronchitis"[All
Fields])) OR ("pulmonary emphysema"[MeSH Terms] OR ("pulmonary"[All Fields] AND
"emphysema"[All Fields]) OR "pulmonary emphysema"[All Fields] OR "emphysema"[All Fields] OR
"emphysema"[MeSH Terms]) OR "Respiratory function tests"[Mesh] OR "pulmonary disease, chronic
obstructive"[Mesh])AND("humans"[MeSHTerms]AND"adult"[MeSHTerms])
141
SearchinEMBASEonAugust4,2012yielding5088hits:
1.expdietaryfiber/
2.fructoseoligosaccharide/ordextrin/orguargum/orpectin/
3.exphighfiberdiet/
4.expfiber/
5.*fiber/
6.cereal/orpolysaccharide/orpectin/
7.expglucan/
8.expINULIN/
9.expDEXTRIN/
10.expLIGNIN/
11.expCHITIN/
12.expoligosaccharide/
13.expARABINOXYLAN/
14.expCELLULOSE/
15.expfattyacid/
16.expsaturatedfattyacid/
17.expunsaturatedfattyacid/
18. exp fish oil/ or exp omega 3 fatty acid/ or exp docosahexaenoic acid/ or exp icosapentaenoic
acid/orexpfatintake/
19.expomega6fattyacid/
20.eicosapentaenoicacid.mp.orexpicosapentaenoicacid/
21.PUFA.mp.
22.SFA.mp.
23.MUFA.mp.
24.EPA.mp.
25.DHA.mp.
26.alpha‐linolenicacid.mp.orexplinolenicacid/
27.explinolenicacid/
28.n‐3fatty.mp.
29.n‐6fatty.mp.
30.exparachidonicacid/
31.expgammalinolenicacid/
32.exptransfattyacid/
33.exppalmitoleicacid/orexphexadecatrienoicacid/
34.expstearidonicacid/
35.eicosatrienoicacid.mp.orexpicosatrienoicacid/
36.eicosatetraenoicacid.mp.orexpicosatetraenoicacid/
37.clupanodonicacid.mp.
38.tetracosapentaenoicacid.mp.
39.eicosadienoicacid.mp.
40.expdihomogammalinolenicacid/
41.docosadienoicacid.mp.
42.expadrenicacid/
43.expdocosapentaenoicacid/
44.expconjugatedlinolenicacid/orexpcalendicacid/orexpconjugatedlinoleicacid/
45.expvolatilefattyacid/orexpmyristoleicacid/
46.explauricacid/orexpsapienicacid/
47.expoleicacid/
48.experucicacid/
49.explauricacid/
50.expmyristicacid/
51.exppalmiticacid/
52.expstearicacid/
53.exparachidicacid/
54.expbehenicacid/
55.explignocericacid/
56.ceroticacid.mp.orexpoleanolicacid/orexphexacosanoicacid/
142
CHAPTER7
57.expverylongchainfattyacid/orexplongchainfattyacid/orexpshortchainfattyacid/orexp
aceticacid/orexpbutyricacid/
58.exp3hydroxybutyricacid/orexpmediumchainfattyacid/
59.expBUTTER/
60.expcoconutoil/
61.exppalmoil/
62.expcottonseedoil/
63.expwheatgermoil/orexpvegetableoil/
64.soyaoil.mp.orexpsoybeanoil/
65.expoliveoil/
66.expcornoil/
67.expsunfloweroil/
68.hempoil.mp.
69.exprapeseedoil/orexpcanolaoil/
70.expFRUIT/
71.expVEGETABLE/
72.expMEAT/
73.expFISH/
74.expCHICKENMEAT/
75.expegg/
76.expnut/
77.1or2or3or4or5or6or7or8or9or10or11or12or13or14or15or16or17or18or19
or20or21or22or23or24or25or26or27or28or29or30or31or32or33or34or35or36
or71or72or73or74or75or76
78.explungfunction/
79.pulmonaryfunction.mp.
80.exprespiratoryfunction/
81.ventilatoryfunction.mp.
82.expchronicbronchitis/orexpwheezing/orrespiratorysymptoms.mp.orexpcoughing/
83.pulmonarysymptoms.mp.
84.breathlessness.mp.orexpdyspnea/
85.expcoughing/orexpsputum/orphlegm.mp.
86.dyspnoea.mp.
87.FEV.mp.orexpforcedexpiratoryvolume/
88. exp vital capacity/ or exp chronic obstructive lung disease/ or exp airway obstruction/ or exp
lungfunctiontest/orFEV1.mp.
89.FVC.mp.
90.expforcedvitalcapacity/
91.COPD.mp.
92.expLUNGEMPHYSEMA/orexpEMPHYSEMA/
93.78or79or80or81or82or83or84or85or86or87or88or89or90or91or92
94.77and93
95.limit94to(humanand(adult<18to64years>oraged<65+years>))
SearchinCochraneCollaborationDatabaseonAugust4,2012yielding218hits:
(dietaryfiberORsolublefiberORinsolublefiberORfibreORnonstarchpolysaccharidesORglucans
OR inulin OR dextrins OR lignin OR waxes OR chitin OR pectins OR oligosaccharides OR gums OR
resistantstarchesORroughageORroughagesORarabinoxylansORcelluloseORfattyacidsORSFA
OR saturated fatty acids OR MUFA OR monounsaturated fatty acids OR PUFA OR polyunsaturated
fatty acids OR n‐3 fatty OR omega‐3 OR omega‐6 OR EPA OR eicosapentaenoic acid OR DHA OR
docosahexaenoic acid OR ALA OR alpha‐linolenic acid OR n‐6 fatty OR linoleic acid OR arachidonic
acidORgamma‐linolenicacidORtransfattyacidsORHexadecatrienoicacidORStearidonicacidOR
EicosatrienoicacidOREicosatetraenoicacidORClupanodonicacidORTetracosapentaenoicacidOR
TetracosahexaenoicacidOREicosadienoicacidORDihomogammalinolenicacidORDocosadienoic
acidORAdrenicacidORDocosapentaenoicacidORCalendicacidORMyristoleicacidORPalmitoleic
143
acidORSapienicacidOROleicacidORerucicacidORlauricacidORmyristicacidORpalmiticacidOR
stearicacidORarachidicacidORbehenicacidORlignocericacidORceroticacidORlardORbutter
ORcoconutoilORpalmoilORcottonseedoilORwheatgermoilORsoyaoilORoliveoilORcornoil
ORsunfloweroilORhempoilORcanolaoilORshortchainfattyacidsORmediumchainfattyacids
ORlongchainfattyacidsORrapeseedoilORfruitsORvegetablesORmeatORfowlORgrainOReggs
ORnutsORfishOR("DietaryFiber"[Mesh])OR("Polysaccharides"[Mesh])OR("Dietaryfats"[Mesh])
OR ("Fatty acids"[Mesh]) OR ("Fish Oils"[Mesh])) AND (lung function OR pulmonary function OR
respiratory function OR respiratory symptoms OR pulmonary symptoms OR breathlessness OR
cough OR phlegm OR dyspnea OR dyspnoea OR sputum OR FEV OR FEV1 OR forced expiratory
volume OR forced vital capacity OR FVC OR COPD OR chronic obstructive pulmonary disease OR
chronic bronchitis OR emphysema OR ("Respiratory function tests"[Mesh]) OR (“Lung diseases,
obstructive"[Mesh])OR("pulmonarydisease,chronicobstructive"[Mesh]))
144
CHAPTER8
Discussion
Central fat and peripheral muscle:
partnersincrimeinCOPD
BramvandenBorst,HarryR.Gosker,AnnemieM.W.J.Schols
American Journal of Respiratory and Critical Care Medicine
2013;187(1):8‐13
145
DISCUSSION
Introduction
According to the World Health Organisation, by the year 2020, chronic
obstructivepulmonarydisease(COPD)isexpectedtobethethird‐leadingcause
of death, and it is the only major chronic disease with an increasing mortality
rate. Successful efforts to reduce this mortality will require a better
understanding of the relationship between risk factors and the severity of the
underlyinglungdisease.Inpatientswithadvanceddisease,respiratoryfailureis
the most common cause of death and weight loss and low muscle mass and
strengthareimportantdeterminantsofmortality[1].Inthesepatients,therisk
ofdeathisremarkablylowerinobesepatients.Incontrast,inpatientswithmild‐
to‐moderate disease, the primary cause of death is ischemic cardiovascular
disease, for which overweight and obesity are important risk factors [2]. The
ongoingworldwideobesityepidemicnecessitatesanenhancedunderstandingof
the interactions between cigarette smoke exposure, cardiovascular disease,
skeletal muscle dysfunction and adiposity in patients with COPD. In this essay,
wewilldiscusssomeofthemechanismslinkingmetabolismandadipositywith
clinicaloutcomesinpatientswithCOPDandoffertailoredlifestyleinterventions
thatmayreducemorbidityandmortalityinthesepatients.
DriversofcardiovascularandmetabolicriskinCOPD
Lossofperipheralskeletalmuscleoxidativephenotype
MuscleatrophyishighlyprevalentinCOPDirrespectiveofdiseaseseverity[3,4].
In moderate‐to‐severe COPD, this atrophy is associated with a reduction in the
activity of oxidative enzymes, a decrease in mitochondrial density, impaired
mitochondrial function, and a fiber type I‐to‐II shift [5]. This loss of skeletal
muscle “oxidative phenotype” has been associated with muscle fatigue [6] and
decreasedmechanicalefficiency[7],andhasrecentlyalsobeendemonstratedin
patientswithmild‐to‐moderateCOPDintheabsenceofmuscleatrophy[8].
In diabetes, it has been proposed that skeletal muscle mitochondrial
dysfunctionmayunderlieperipheralinsulinresistanceviacomplexmechanisms
including oxidative stress, intramuscular lipid accumulation, and inflammation
[9, 10]. However, recent tests of this hypothesis in human subjects have
suggested that the mitochondrial defect observed in the skeletal muscles of
patientswithdiabetesmightbesecondarytotheinsulin‐resistantstateinstead
ofbeingacausalfactor(reviewedin[11]).Thisnotionissupportedbystriking
parallels in skeletal muscle oxidative phenotype between underweight/normal
weightnon‐diabeticCOPDpatientsandoverweight/obesediabetespatients.For
example,decreasedmitochondrialrespirationhasalsobeenreportedinpatients
withtype2diabetes[12,13],andthemagnitudeofreducedskeletalmusclePGC‐
1α gene (the master regulator of mitochondrial biogenesis) expression is
146
CHAPTER8
comparable between non‐obese COPD patients [8, 14] and obese subjects with
type2diabetes[15].Also,severalstudieshaveshownevidenceforadecreased
typeIfiberproportioninoverweight/obesetype2diabetespatientscompared
to BMI‐matched controls [16, 17]. The decreased oxidative capacity of the
skeletalmusclehasbeensuggestedtoplayamajorrolein‘metabolicinflexibility’
[18],whichisdefinedbyareducedcapacitytoincreasefatoxidationinresponse
toincreasedfattyacidavailabilityandanimpairedswitchfromfattoglucoseas
the primary fuel source after a meal. Metabolic inflexibility may lead to the
accumulation of intramyocellular lipids which has been associated with insulin
resistance through mechanisms possibly involving the lipid intermediates
ceramidesanddiacylglycerolthatinterferewithinsulinsignalingpathways[19].
In case of high caloric intake and lack of physical exercise, a positive energy
balancemaythenfurtherenhanceectopicfatdeposition.
Investigators have employed homeostatic modeling of insulin resistance
(HOMA index) to show that insulin sensitivity is lower in non‐obese and non‐
diabetic COPD patients compared with BMI‐matched healthy controls [20‐22].
The HOMA index, however, primarily reflects hepatic insulin sensitivity, while
skeletal muscle insulin sensitivity can be most accurately assessed with the
euglycaemic insulin clamp technique. Therefore, clamp studies are indicated in
ordertoproofordisputethatadecreasedskeletalmuscleoxidativephenotypeis
associatedwithinsulinresistanceatthelevelofskeletalmuscleinCOPDasitis
indiabetes.
Visceralobesity
Several groups of investigators have suggested that low‐grade systemic
inflammationcharacteristicofanumberofdiseasesincludingCOPDisassociated
with an increased risk for the development of atherosclerosis and ischemic
cardiovascularevents[23].Thereisincreasingevidencethatadiposetissueisa
significantcontributortothesystemicinflammatoryloadinCOPD.Forexample,
apersistentsystemicinflammationphenotypewasrecentlydescribedinpatients
withCOPDandwasassociatedwithbodymassindex(BMI)butnotwithfat‐free
mass index, suggesting an inflammatoryrole foradipose tissue [24]. Moreover,
in a subgroup of COPD patients, low‐grade systemic inflammation has been
positively related to total abdominal fat mass [25]. Thus, adipose tissue
dysfunction,includingenhancedadiposetissueinflammation,maycontributeto
low‐gradesystemicinflammationinCOPDpatientswithabsoluteorrelativefat
abundance.
Abdominal visceral fat appears to be more strongly associated than
subcutaneous fat with risk factors for cardiovascular and metabolic disorders,
including hypertension and insulin resistance [26, 27], and is receiving
increasingattentioninCOPD.Indeed,tworecentstudieshavereportedexcessive
visceral fat mass in non‐obese mild‐to‐moderate COPD patients as detected by
147
DISCUSSION
CT scanning [28, 29]. Moreover, excessive visceral fat mass was positively
associated with circulating interleukin (IL)‐6 levels, which were strongly
associatedwithall‐causeandCVDmortality[28].Importantly,excessivevisceral
fat mass in these COPD patients existed in the absence of obesity or increased
abdominal circumference compared with the control persons in these studies,
suggestingenhancedfataccumulationspecificallyinthevisceralcompartment.It
isunclearwhetherthepulmonaryimpairmentperseoranoverallpoorlifestyle,
or both, predispose to excessive visceral fat accumulation. A similar selective
increase in visceral fat mass has been observed in other chronic diseases
characterized by tissue inflammation including rheumatoid arthritis, Crohn’s
disease, and psoriasis. These intriguing findings suggest a link between
inflammation, visceral fat and cardiovascular risk. Careful mechanistic studies
willberequiredtodeterminewhetheranincreaseinvisceralfatplaysacausal
roleinthischain.
Theinflammatorycapacityofvisceralfatisconsiderablygreaterthanthatof
subcutaneous fat, and it has been suggested that visceral fat is an important
source of IL‐6 and IL‐1β [30]. In addition, there is evidence that an increase in
visceral relative to subcutaneous fat is associated with hypertriglyceridaemia
and a decreased rate of glucose utilization in obese humans [31]. Differential
expressionofthefatmassandobesityassociated(FTO)genehasbeenassociated
with the differential deposition of fat in the visceral or subcutaneous stores,
adipose tissue inflammation and differences in the expression of inflammatory
and insulin resistance‐related genes in the visceral compared with the
subcutaneous fat in otherwise healthy overweight/obese subjects [32].
Consistentwiththehypothesisthatthisgenemaycontributetoobesity‐related
inflammation in patients with COPD, single nucleotide polymorphisms in FTO
genewerereportedtobepositivelyassociatedwithBMIinpatientswithCOPD[33].
Increased visceral fat mass has been associated with increased portal
drainageoffreefattyacidsandadipokinestotheliver,whichmayinducehepatic
insulin resistance, inflammation and oxidative stress (‘portal hypothesis’ [34]).
Furthermore,visceraladiposityhasbeenassociatedwithfatdepositioninother
undesirable sites such as the liver, the heart, and the skeletal muscle [35]. To
date,thereisnodataavailableifectopicfatdepositionplaysaroleintheriskof
cardiovascular and metabolic disease in COPD, but it may be involved in those
withincreasedvisceralfatmass.
Thereisincreasingevidencethatcomponentsofthemetabolicsyndromeare
more prevalent in overweight patients with mild‐to‐moderate COPD compared
withBMI‐matchedcontrols.AstudyreportedthatinoverweightCOPDpatients
47% had the metabolic syndrome compared to 21% in BMI‐matched healthy
subjects[36].Inanotherstudyof170GermanpatientswithCOPD,53%ofGOLD
IIpatientsmetthecriteriaformetabolicsyndromeandasmuchas83%fulfilled
the criteria for abdominal obesity [37]. Congruently, a large population study
showed that those with airflow obstruction had greater odds of having the
148
CHAPTER8
metabolic syndrome when adjusted for BMI, and when its components were
further analyzed only abdominal obesity was significantly related to airflow
obstruction[38].
Lifestyleinterventiontoreducetheriskofcardiovascularand
metabolicdiseaseinpatientswithCOPD
Smokingcessation
It is well‐known that smoking isan independent riskfactor for both COPDand
cardiovascular disease. Less recognized is that smokers have more central
adiposity[39]andexposuretosmokingisassociatedwithpoorskeletalmuscle
function [40]. Very recent data showed that smokers are characterized by
decreased skeletal muscle insulin sensitivity, which was partially reversible
uponsmokingcessation.Experimentalevidencesuggestsinvolvementofskeletal
muscle mammalian target of rapamycin (mTOR) in this process, a crucial
regulator of cellular protein synthesis [41]. BMI, as well as whole‐body muscle
and fat masses, are comparably reduced in older mild‐to‐moderate COPD
patients and age‐matched non‐COPD smokers compared with non‐smokers,
suggesting a common insult earlier in life related to smoking [4]. Furthermore,
the peripheral skeletal muscle of non‐COPD smokers shows a loss of oxidative
phenotype compared to non‐smoking healthy controls [42]. Smoking induces
low‐grade systemic inflammation, independently of COPD [43], and the
persistentsystemicinflammationobservedinpatientswithCOPDwaspredicted
by current smoking status [24]. Notably, chronic exposure of mice to cigarette
smoke also induces systemic inflammation and decreased skeletal muscle
oxidative capacity [44]. Smoking cessation is associated with fat mass gain,
particularly in the visceral compartment [45], stressing the need for broad
lifestyleinterventionsbeyondsmokingcessation.
Physicalactivityandaerobicexercisetraining
The Centers for Disease Control and Prevention and the American College of
Sports Medicine recommend that a person should engage in ≥30 minutes of
moderate‐to‐vigorous physical activity each day, preferably in bouts of ≥10
minutes[46].ThemajorityofpatientswithCOPDhaveasedentarylifestyle[5]
and physical inactivity contributes to the risk and mortality associated with
cardiovascular and metabolic disease. Patients with COPD, irrespective of
diseaseseverity,havefewer,shorterandlessintenseperiodsofintensephysical
activity,andspendmoretimesedentarycomparedwithhealthysubjects[8].It
has been argued that the reduced level of physical activity level is causally
implicated in the development of COPD, as studies have consistently reported
149
DISCUSSION
attenuatedFEV1declineinphysicallyactivepersons(smokersandnon‐smokers)
[47].
Bed‐reststudiesinhealthy,leansubjectshaveshownthatphysicalinactivity
induces a wide array of metabolic abnormalities including skeletal muscle
insulin resistance, impaired skeletal muscle lipid trafficking, a shift in fuel
metabolism in favor of carbohydrate oxidation and in detriment of lipid
oxidation, an increase in intramyocellular lipid content, muscle atrophy and
muscle fiber type I‐to‐II shift, and increased plasma non‐esterified fatty acid
levels [48]. As many of these metabolic abnormalities have also been found in
COPD,theadverseeffectsofphysicalinactivityarelikelytobesubstantial.
The benefits of aerobic exercise training on cardiometabolic health have
been well‐described in overweight or obese adults with and without type 2
diabetes,buthavebeenincompletelystudiedinpatientswithCOPD.Inhealthy
overweight or obese adults, aerobic exercise training improves insulin
sensitivityandinducesmitochondrialbiogenesisintheskeletalmuscle[49,50],
andinducesthelossofvisceralfatmass[51].Theseeffectswereobservedeven
inpatientswhosebodymassorwaistcircumferencedidnotchange.
Classically, it was assumed that aerobic exercise training intensity in COPD
could not reach the levels required to induce physiological effects due to
ventilatory limitation. Two important papers disputed this assumption by
showing improved lactate‐to‐work ratios and increased activity of key
mitochondrial enzymes in muscle biopsies after aerobic exercise training [52,
53]andthesefindingshavebeensubsequentlyconfirmed.
Arterialstiffnessisanindependentcardiovascularriskfactor,whichisalso
improvedbyaerobicexercisetraininginhealthysubjects[54].Increasedarterial
stiffness has been reported in patients with COPD, particularly those with
advanced disease [55]. Interestingly, Vivodtzev et al. reported that a relatively
short aerobic exercise training program improved arterial stiffness and
quadricepsmuscleenduranceinacohortofpatientswithCOPD[56].Galeetal.
[57]showedthatstandardmultidisciplinarypulmonaryrehabilitationincluding
aerobic exercise training reduced arterial stiffness without changing BMI in
mildlyoverweightCOPDpatients.
Dietandnutritionalmodulation
While weight gain and improved nutrition are recommended for underweight
patientswithCOPD[58],thereisnosystematicstudydemonstratingefficacyof
improved nutrition or weight reduction strategies in overweight or obese
patients with COPD. It seems likely that the obesity paradox has limited
enthusiasm for this approach. However, data from weight loss interventions in
patientsatriskfordiabetessuggestthatevenmodestreductionsinweightcan
reduce the risk of cardiovascular and metabolic diseases, perhaps through
improvementsinbodyfatdistribution.Forexample,an~2‐5%lossofbodymass
150
CHAPTER8
wasreportedtobeassociatedwithasustainedandpreferentiallossofvisceral
relativetosubcutaneousfatinotherwisehealthyoverweightorobesepersons,
however, weight loss >5% was associated with the loss of fat in both
compartments[59].
Dietary factors have recently also been put forward as contributing
environmental factors in COPD etiology and pathology. A diet rich in fruit,
vegetables,whole‐mealcerealsandfishisassociatedwithareducedCOPDrisk,
whereas a “western diet” rich in refined grains, cured and red meats, desserts
andFrenchfriesisassociatedwithincreasedCOPDrisk,independentofsmoking
[60, 61]. Further study is required to identify the specific dietary components
withCOPDriskasthesestrategiesmightbeusedtopreventthedevelopmentor
slowtheprogressionofCOPDwhilesimultaneouslyreducingcardiovascularrisk.
The available literature suggests a particular important role for dietary fiber.
ThereisconvincingevidenceoflowerfiberintakeinCOPDpatients[28]andthat
increaseddietaryfiberintakeisinverselyassociatedwithrespiratorymortality
[62]. Dietary fiber has been shown to alter gut immunity and reduce systemic
inflammation, which may be mediated by alterations in the gut microbiome or
other mechanisms. Poly‐unsaturated fatty acids (PUFAs) warrant further
investigationintargetednutritionalmodulationasPUFAs(inparticularn‐3fatty
acids) are known agonists of peroxisome proliferator‐activated receptors
(PPARs), which are implicated in skeletal muscle oxidative gene expression.
PUFA supplementation was also shown to enhance training effects on exercise
capacityinarandomizedcontrolledtrialthatincludedpatientswithmoderate‐
to‐severe COPD [63]. PPARs as potential targets for nutritional as well as
pharmacological modulation in COPD have been discussed in detail elsewhere
[64].
Another promising example of nutritional modulation to improve the
cardiovascular and metabolic disease risk profile in COPD is resveratrol, a
naturalpolyphenoliccompoundinnutsandgrapesthatiswidelyavailableasa
nutritional supplement. Resveratrol supplementation (150 mg/day) has
emerged as a calorie restriction mimetic with beneficial cardiovascular, anti‐
agingandanti‐diabetogenicpropertiesinotherwisehealthyoverweightorobese
men [65]. Interestingly, pharmacological inhibition of phosphodiesterase‐4
reproduced all of the metabolic benefits of resveratrol in experimental studies,
likely through the same potential mechanisms. These include elevation of
intracellular cAMP levels, thereby increasing NAD(+) and the activity of SirT1,
whichinconcertprovideadriveforincreasingoxidativemetabolism[66].
Concludingremarks
Adversemetabolichealth,intermsoflossofperipheralskeletalmuscleoxidative
phenotype and excessive visceralfat accumulation isassociated with increased
morbidityandmortalityinpatientswithCOPD.Thisincreasedriskisevidentat
151
DISCUSSION
all stages of the disease and is not necessarily restricted to obese patients.
Rudermanetal.describedthisphenotypeasa“metabolicallyobese,butnormal
weight person” [67]. Physical inactivity, poor dietary quality, and smoking
contributetothesemetabolicabnormalities,andinconcertfurtherincreasethe
riskofcardiovascularmorbidityandmortality(Figure1).
Because loss of muscle oxidative phenotype and excessive visceral fat
accumulation can exist without changes in muscle mass and abdominal
circumference,thereisaneedfornewdiagnostictoolsandbiomarkerstoreveal
these hidden phenomena. Quadriceps muscle endurance tests or the lactate
responseuponexercisemayprovidegoodindicatorsofskeletalmuscleoxidative
phenotype, and abdominal bioelectrical impedance, ultrasound and plasma
levels of the pro‐thrombotic adipokine plasminogen activator inhibitor‐1 have
shown good correlations with CT‐acquired visceral fat mass but require
validationinCOPD.
Themostimportantquestionthatneedstobeansweredinfuturestudiesis
whethertheadversemetabolicprofileinCOPDpatientsiscausedbythedisease
itself or by an overall poor lifestyle, and whether the two act synergistically.
Studies in patients with COPD and control subjects without COPD matched for
overall lifestyle may help in disentangling COPD‐specific versus lifestyle‐
mediatedinfluences.Otherimportantquestionsremain:Isthelossofperipheral
skeletal muscle oxidative phenotype in COPD an accelerator of or causally
implicatedininsulinresistance?Isthereaphysiologicallinkbetweenthelossof
peripheralskeletalmuscleoxidativephenotypeandvisceralfataccumulation?Is
theinflammatorystatusofvisceralversussubcutaneousfatenhancedinCOPD,
anddoesthis relate toanadverse cardiovascular ormetabolic profile? What is
the role of ectopic fat deposition in the adverse cardiovascular and metabolic
profile in COPD? Finally, a possible link between adipose hormones and
pulmonary inflammation may shed new lights on the importance of adipose
tissueinCOPDprogressioninthecurrentobesogenicsociety.
Itisprojectedthatincreasedknowledgeofthemetabolicchangesinpatients
with COPD will allow investigators to test interventions aimed at reducing the
riskofcardiovascularandmetabolicdiseasesandtheirassociatedmorbidityand
mortality. Available data suggest that the early implementation of tailored
lifestyleinterventionsinpatientswithCOPDwhoarenotyetphysicallyimpaired
are likely to be beneficial. These lifestyle factors extend beyond
recommendations for smoking cessation, instead an integrated systems
approach addressing smoking, physical inactivity, and poor diet may be most
beneficialinslowingtheprogressionofthedisease.
152
CHAPTER8
Figure 1. Conceptual representation of the drivers of cardiovascular and
metabolicriskinCOPD.
153
DISCUSSION
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
154
SwallowEB,ReyesD,HopkinsonNS,ManWD,PorcherR,CettiEJ,MooreAJ,MoxhamJ,PolkeyMI.Quadriceps
strength predicts mortality in patients with moderate to severe chronic obstructive pulmonary disease.
Thorax.2007;62(2):115‐20.
McGarveyLP,JohnM,AndersonJA,ZvarichM,WiseRA.Ascertainmentofcause‐specificmortalityinCOPD:
operationsoftheTORCHClinicalEndpointCommittee.Thorax.2007;62(5):411‐5.
ShrikrishnaD,PatelM,TannerRJ,SeymourJM,ConnollyBA,PuthuchearyZA,WalshSL,BlochSA,SidhuPS,
Hart N, et al. Quadriceps wasting and physical inactivity in patients with COPD. Eur Respir J. 2012 Feb 23
[Epubaheadofprint].
vandenBorstB,KosterA,YuB,GoskerHR,MeibohmB,BauerDC,KritchevskySB,LiuY,NewmanAB,Harris
TB,etal.Isage‐relateddeclineinleanmassandphysicalfunctionacceleratedbyobstructivelungdiseaseor
smoking?Thorax.2011;66(11):961‐9.
Rabinovich RA, Vilaro J. Structural and functional changes of peripheral muscles in chronic obstructive
pulmonarydiseasepatients.CurrOpinPulmMed.2010;16(2):123‐33.
LayecG,HaselerLJ,RichardsonRS.TheeffectofhigherATPcostofcontractiononthemetabolicresponseto
gradedexerciseinpatientswithchronicobstructivepulmonarydisease.JApplPhysiol.2012;112(6):1041‐8.
vandenBorstB,SlotIGM,HellwigVACV,VosseBAH,KeldersMCJM,BarreiroE,ScholsAMWJ,GoskerHR.Loss
ofquadricepsmuscleoxidativephenotypeanddecreasedenduranceinpatientswithmild‐to‐moderateCOPD.
JournalofAppliedPhysiology.2012Jul19[Epubaheadofprint].
Schrauwen P, Schrauwen‐Hinderling V, Hoeks J, Hesselink MK. Mitochondrial dysfunction and lipotoxicity.
BiochimBiophysActa.2010;1801(3):266‐71.
WangCH,WangCC,WeiYH.Mitochondrialdysfunctionininsulininsensitivity:implicationofmitochondrial
roleintype2diabetes.AnnNYAcadSci.2010;1201:157‐65.
HoeksJ,SchrauwenP.Musclemitochondriaandinsulinresistance:ahumanperspective.TrendsEndocrinol
Metab.2012;23(9):444‐50.
Mogensen M, Sahlin K, Fernstrom M, Glintborg D, Vind BF, Beck‐Nielsen H, Hojlund K. Mitochondrial
respirationisdecreasedinskeletalmuscleofpatientswithtype2diabetes.Diabetes.2007;56(6):1592‐9.
RabolR,LarsenS,HojbergPM,AlmdalT,BoushelR,HaugaardSB,AndersenJL,MadsbadS,DelaF.Regional
anatomic differences in skeletal muscle mitochondrial respiration in type 2 diabetes and obesity. J Clin
EndocrinolMetab.2010;95(2):857‐63.
Remels AH, Schrauwen P, Broekhuizen R, Willems J, Kersten S, Gosker HR, Schols AM. Peroxisome
proliferator‐activated receptor expression is reduced in skeletal muscle in COPD. Eur Respir J.
2007;30(2):245‐52.
Mensink M, Hesselink MK, Russell AP, Schaart G, Sels JP, Schrauwen P. Improved skeletal muscle oxidative
enzyme activity and restoration of PGC‐1 alpha and PPAR beta/delta gene expression upon rosiglitazone
treatmentinobesepatientswithtype2diabetesmellitus.IntJObes(Lond).2007;31(8):1302‐10.
SegerstromAB,ElgzyriT,ErikssonKF,GroopL,ThorssonO,WollmerP.Exercisecapacityinrelationtobody
fatdistributionandmusclefibredistributioninelderlymalesubjectswithimpairedglucosetolerance,type2
diabetesandmatchedcontrols.DiabetesResClinPract.2011;94(1):57‐63.
MarinP,AnderssonB,KrotkiewskiM,BjorntorpP.Musclefibercompositionandcapillarydensityinwomen
andmenwithNIDDM.DiabetesCare.1994;17(5):382‐6.
Kelley DE, Mandarino LJ. Fuel selection in human skeletal muscle in insulin resistance: a reexamination.
Diabetes.2000;49(5):677‐83.
Coen PM, Goodpaster BH. Role of intramyocellular lipids in human health. Trends Endocrinol Metab.
2012;23(8):391‐8.
van den Borst B, Gosker HR, Wesseling G, de Jager W, Hellwig VA, Snepvangers FJ, Schols AM. Low‐grade
adiposetissueinflammationinpatientswithmild‐to‐moderatechronicobstructivepulmonarydisease.AmJ
ClinNutr.2011;94(6):1504‐12.
BoltonCE,EvansM,IonescuAA,EdwardsSM,MorrisRH,DunseathG,LuzioSD,OwensDR,ShaleDJ.Insulin
resistanceandinflammation‐AfurthersystemiccomplicationofCOPD.COPD.2007;4(2):121‐6.
FranssenFM,SauerweinHP,AckermansMT,RuttenEP,WoutersEF,ScholsAM.Increasedpostabsorptiveand
exercise‐induced whole‐body glucose production in patients with chronic obstructive pulmonary disease.
Metabolism.2011;60(7):957‐64.
Sin DD, Man SF. Why are patients with chronic obstructive pulmonary disease at increased risk of
cardiovascular diseases? The potential role of systemic inflammation in chronic obstructive pulmonary
disease.Circulation.2003;107(11):1514‐9.
Agusti A, Edwards LD, Rennard SI, Macnee W, Tal‐Singer R, Miller BE, Vestbo J, Lomas DA, Calverley PM,
Wouters E, et al. Persistent Systemic Inflammation is Associated with Poor Clinical Outcomes in COPD: A
NovelPhenotype.PLoSOne.2012;7(5):e37483.
Rutten EP, Breyer MK, Spruit MA, Hofstra T, van Melick PP, Schols AM, Wouters EF. Abdominal fat mass
contributes to the systemic inflammation in chronic obstructive pulmonary disease. Clin Nutr.
2010;29(6):756‐60.
KishidaK,FunahashiT,MatsuzawaY,ShimomuraI.Visceraladiposityasatargetforthemanagementofthe
metabolicsyndrome.AnnMed.2012;44(3):233‐41.
CHAPTER8
27. PerriniS,LeonardiniA,LaviolaL,GiorginoF.Biologicalspecificityofvisceraladiposetissueandtherapeutic
intervention.ArchPhysiolBiochem.2008;114(4):277‐86.
28. vandenBorstB,GoskerHR,KosterA,YuB,KritchevskySB,LiuY,MeibohmB,RiceTB,ShlipakM,YendeS,et
29. FurutateR,IshiiT,WakabayashiR,MotegiT,YamadaK,GemmaA,KidaK.Excessivevisceralfataccumulation
inadvancedchronicobstructivepulmonarydisease.IntJChronObstructPulmonDis.2011;6:423‐30.
31. Klimcakova E, Roussel B, Kovacova Z, Kovacikova M, Siklova‐Vitkova M, Combes M, Hejnova J, Decaunes P,
Maoret JJ, Vedral T, et al. Macrophage gene expression is related to obesity and the metabolic syndrome in
humansubcutaneousfataswellasinvisceralfat.Diabetologia.2011;54(4):876‐87.
32. SamarasK,BotelhoNK,ChisholmDJ,LordRV.SubcutaneousandvisceraladiposetissueFTOgeneexpression
and adiposity, insulin action, glucose metabolism, and inflammatory adipokines in type 2 diabetes mellitus
andinhealth.ObesSurg.2010;20(1):108‐13.
33. WanES,ChoMH,BoutaouiN,KlandermanBJ,SylviaJS,ZinitiJP,WonS,LangeC,PillaiSG,AndersonWH,etal.
Genome‐wide association analysis of body mass inchronic obstructive pulmonary disease. AmJRespir Cell
MolBiol.2011;45(2):304‐10.
34. Bjorntorp P. "Portal" adipose tissue as a generator of risk factors for cardiovascular disease and diabetes.
Arteriosclerosis.1990;10(4):493‐6.
35. DespresJP,LemieuxI.Abdominalobesityandmetabolicsyndrome.Nature.2006;444(7121):881‐7.
36. MarquisK,MaltaisF,DuguayV,BezeauAM,LeBlancP,JobinJ,PoirierP.Themetabolicsyndromeinpatients
withchronicobstructivepulmonarydisease.JCardiopulmRehabil.2005;25(4):226‐32;discussion33‐4.
37. Watz H, Waschki B, Kirsten A, Muller KC, Kretschmar G, Meyer T, Holz O, Magnussen H. The metabolic
syndromeinpatientswithchronicbronchitisandCOPD:frequencyandassociatedconsequencesforsystemic
inflammationandphysicalinactivity.Chest.2009;136(4):1039‐46.
38. Lam KB, Jordan RE, Jiang CQ, Thomas GN, Miller MR, Zhang WS, Lam TH, Cheng KK, Adab P. Airflow
obstructionandmetabolicsyndrome:theGuangzhouBiobankCohortStudy.EurRespirJ.2010;35(2):317‐23.
39. CanoyD,WarehamN,LubenR,WelchA,BinghamS,DayN,KhawKT.Cigarettesmokingandfatdistributionin
21,828Britishmenandwomen:apopulation‐basedstudy.ObesRes.2005;13(8):1466‐75.
40. Strand BH, Mishra G, Kuh D, Guralnik JM, Patel KV. Smoking history and physical performance in midlife:
resultsfromtheBritish1946birthcohort.JGerontolABiolSciMedSci.2011;66(1):142‐9.
41. Bergman BC, Perreault L, Hunerdosse D, Kerege A, Playdon M, Samek AM, Eckel RH. Novel and Reversible
MechanismsofSmoking‐InducedInsulinResistanceinHumans.Diabetes.2012;61(12):3156‐66.
42. Montes de Oca M,Loeb E, Torres SH, De SanctisJ, Hernandez N, Talamo C. Peripheral muscle alterationsin
non‐COPDsmokers.Chest.2008;133(1):13‐8.
43. Yanbaeva DG, Dentener MA, Creutzberg EC, Wesseling G, Wouters EF. Systemic effects of smoking. Chest.
2007;131(5):1557‐66.
44. Gosker HR, Langen RC, Bracke KR, Joos GF, Brusselle GG, Steele C, Ward KA, Wouters EF, Schols AM.
Extrapulmonary manifestations of chronic obstructive pulmonary disease in a mouse model of chronic
cigarettesmokeexposure.AmJRespirCellMolBiol.2009;40(6):710‐6.
45. MatsushitaY,NakagawaT,YamamotoS,TakahashiY,NodaM,MizoueT.Associationsofsmokingcessation
withvisceralfatareaandprevalenceofmetabolicsyndromeinmen:theHitachihealthstudy.Obesity(Silver
Spring).2011;19(3):647‐51.
46. PateRR,PrattM,BlairSN,HaskellWL,MaceraCA,BouchardC,BuchnerD,EttingerW,HeathGW,KingAC,et
47. Hopkinson NS, Polkey MI. Does physical inactivity cause chronic obstructive pulmonary disease? Clin Sci
(Lond).2010;118(9):565‐72.
48. Bergouignan A, Rudwill F, Simon C, Blanc S. Physical inactivity as the culprit of metabolic inflexibility:
evidencefrombed‐reststudies.JApplPhysiol.2011;111(4):1201‐10.
49. Olesen J, Kiilerich K, Pilegaard H. PGC‐1alpha‐mediated adaptations in skeletal muscle. Pflugers Arch.
2010;460(1):153‐62.
50. Bruce CR, Hawley JA. Improvements in insulin resistance with aerobic exercise training: a lipocentric
approach.MedSciSportsExerc.2004;36(7):1196‐201.
51. IsmailI,KeatingSE,BakerMK,JohnsonNA.Asystematicreviewandmeta‐analysisoftheeffectofaerobicvs.
resistanceexercisetrainingonvisceralfat.ObesRev.2012;13(1):68‐91.
52. CasaburiR,PatessioA,IoliF,ZanaboniS,DonnerCF,WassermanK.Reductionsinexerciselacticacidosisand
ventilation as a result of exercise training in patients with obstructive lung disease. Am Rev Respir Dis.
1991;143(1):9‐18.
53. MaltaisF,LeBlancP,SimardC,JobinJ,BerubeC,BruneauJ,CarrierL,BelleauR.Skeletalmuscleadaptationto
endurance training in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med.
1996;154(2Pt1):442‐7.
54. Mattace‐RasoFU,vanderCammenTJ,HofmanA,vanPopeleNM,BosML,SchalekampMA,AsmarR,Reneman
RS, Hoeks AP, Breteler MM, et al. Arterial stiffness and risk of coronary heart disease and stroke: the
RotterdamStudy.Circulation.2006;113(5):657‐63.
155
DISCUSSION
55. McAllisterDA,MaclayJD,MillsNL,MairG,MillerJ,AndersonD,NewbyDE,MurchisonJT,MacneeW.Arterial
stiffnessisindependentlyassociatedwithemphysemaseverityinpatientswithchronicobstructivepulmonary
disease.AmJRespirCritCareMed.2007;176(12):1208‐14.
56. Vivodtzev I, Minet C, Wuyam B, Borel JC, Vottero G, Monneret D, Baguet JP, Levy P, Pepin JL. Significant
improvementinarterialstiffnessafterendurancetraininginpatientswithCOPD.Chest.2010;137(3):585‐92.
57. Gale NS, Duckers JM, Enright S, Cockcroft JR, Shale DJ, Bolton CE. Does pulmonary rehabilitation address
cardiovascularriskfactorsinpatientswithCOPD?BMCPulmMed.2011;11:20.
58. Schols AM, Slangen J, Volovics L, Wouters EF. Weight loss is a reversible factor in the prognosis of chronic
59. ChastonTB,DixonJB.Factorsassociatedwithpercentchangeinvisceralversussubcutaneousabdominalfat
duringweightloss:findingsfromasystematicreview.IntJObes(Lond).2008;32(4):619‐28.
60. Varraso R, Fung TT, Hu FB, Willett W, Camargo CA. Prospective study of dietary patterns and chronic
61. ShaheenSO,JamesonKA,SyddallHE,SayerAA,DennisonEM,CooperC,RobinsonSM.Therelationofdietary
patternswithadultlungfunctionandCOPD.EurRespirJ.2010;36(2):277‐84.
62. ChuangSC,NoratT,MurphyN,OlsenA,TjonnelandA,OvervadK,Boutron‐RuaultMC,PerquierF,DartoisL,
KaaksR,etal.Fiberintakeandtotalandcause‐specificmortalityintheEuropeanProspectiveInvestigation
intoCancerandNutritioncohort.AmJClinNutr.2012;96(1):164‐74.
63. Broekhuizen R, Wouters EF, Creutzberg EC, Weling‐Scheepers CA, Schols AM. Polyunsaturated fatty acids
improveexercisecapacityinchronicobstructivepulmonarydisease.Thorax.2005;60(5):376‐82.
64. Remels AH, Gosker HR, Schrauwen P, Langen RC, Schols AM. Peroxisome proliferator‐activated receptors: a
therapeutictargetinCOPD?EurRespirJ.2008;31(3):502‐8.
65. TimmersS,KoningsE,BiletL,HoutkooperRH,vandeWeijerT,GoossensGH,HoeksJ,vanderKriekenS,Ryu
D, Kersten S, et al. Calorie restriction‐like effects of 30 days of resveratrol supplementation on energy
metabolismandmetabolicprofileinobesehumans.CellMetab.2011;14(5):612‐22.
66. ParkSJ,AhmadF,PhilpA,BaarK,WilliamsT,LuoH,KeH,RehmannH,TaussigR,BrownAL,etal.Resveratrol
ameliorates aging‐related metabolic phenotypes by inhibiting cAMP phosphodiesterases. Cell.
2012;148(3):421‐33.
67. Ruderman NB, Schneider SH, Berchtold P. The "metabolically‐obese," normal‐weight individual. Am J Clin
Nutr.1981;34(8):1617‐21.
156
Summary
157
SUMMARY
Chronic obstructive pulmonary disease (COPD) affects 64 million people
worldwide and is projected to be the third leading cause of death by the year
2020. Patients with advanced COPD primarily die from respiratory failure,
whereasthosewithmild‐to‐moderateCOPDmostoftendiefromcardiovascular
events. Morbidity and mortality risk in COPD are likely determined by a
combination of disease‐specific factors and an aging and obesogenic society in
general. This thesis has a focus on patients with mild‐to‐moderate COPD and
investigateswhethertheyarecharacterizedbyanincreasedmetabolicriskand
towhatextentagingandanunhealthylifestyleareinvolved.
Peripheral muscle in mild‐to‐moderate COPD: smoked and
tired
Thelossofmusclemassduringtheagingprocessisaphysiologicalphenomenon
(termed ‘sarcopenia’) and has been associated with deterioration in physical
functioning. Many patients with COPD experience muscle weakness that
hampers their daily life activities and the question arises whether this can be
attributedtotheirdiseaseorsimplybeexplainedbysarcopenia.InChapter2it
isshownthattheratewithwhichthesarcopenicprocesstakesplaceinpatients
with COPD is comparable to that of age‐matched subjects without COPD.
However, patients with COPD enter old‐age with much lower absolute muscle
massthansubjectswithoutCOPD.Thus,eitherthedeteriorationofmusclemass
inpatientswithCOPDhadinitiatedearlierinlife,ortheyhadbeenleanallalong.
The Chapter also shows that smokers without COPD are just as lean as the
patients with COPD when entering old‐age. Being the major cause of COPD,
exposure to cigarette smoke thus also seems to affect muscle mass, apparently
earlierinlife.
The mechanisms leadingto muscle dysfunction have been particularly well
studiedinpatientswithsevereCOPD,butnotinpatientswithmilderdegreesof
airflowlimitation.Chapter3providesevidenceforreducedquadricepsmuscle
endurance in patients with mild‐to‐moderate COPD who had normal muscle
mass.Musclefatiguewasassociatedwithaclearlossofoxidativephenotypeof
thequadricepsmusclethatwascharacterizedbyalossofoxidativetypeImuscle
fibers and decreased peroxisome‐proliferator activated receptor‐γ co‐activator
1αgeneexpression(beingthemajorregulatorofoxidativemetabolism).Thus,a
lossofperipheralmuscleoxidativephenotypecanexistalreadyintheabsenceof
musclewasting,andalreadyinpatientswithless‐advancedCOPD.
AdiposetissueinflammationinCOPD:corebusiness
Over the past decade, emerging research has shown that adipose tissue is not
just an energy storage compartment but actively secretes proteins with
158
SUMMARY
inflammatoryproperties,commonlyknownasadipokines.Ithasbeensuggested
thatadysregulationofadipokinesecretionmayplayaroleinsystemicaswellas
pulmonary inflammation in COPD. Chapter 4 presents the first comparison of
adipose tissue inflammatory status between mild‐to‐moderate COPD patients
andmatchedhealthysubjects.Neitheradiposetissueadipokinegeneexpression,
nor adipose tissue macrophage infiltration, nor plasma concentrations of
adipokinesweredifferentbetweentheCOPDpatientsandthehealthysubjects.
Adipositywaspositivelyassociatedwithadiposetissueinflammationinboththe
COPDpatientsandthehealthysubjects.Aspartofmetabolicriskassessment,the
COPD patients did have lower insulin sensitivity compared with the healthy
controls. These data suggest that abdominal subcutaneous adipose tissue
inflammatorystatusisgenerallynotaffectedinmild‐to‐moderateCOPD.
Chapter 5 shows that COPD patients with mild airflow obstruction are
characterized by increased intra‐abdominal fat mass which is not necessarily
associated with obesity or increased abdominal circumference. Plasma
concentrations of interleukin (IL)‐6 were also increased in these patients and
were positively associated with intra‐abdominal fat mass. Also, plasma IL‐6
concentration predicted excess 9‐year mortality of the COPD patients. The
results suggest that interventions aimed at lowering intra‐abdominal fat mass
maybebeneficialinCOPDpatientsintermsofloweringsystemicinflammation
andrelatedmortality.
Experimental studies have suggested that local tissue hypoxia in obesity
inducesadiposetissueinflammation.Ifhypoxiaisindeedoneofthemechanisms
involved,thismaybeofimportanceforCOPDasitoftenassociatedwithhypoxia.
In Chapter 6 it was explored whether mice exposed to chronic hypoxia are
characterized by adipose tissue inflammation. Interestingly, decreased rather
than increased inflammation was found in both subcutaneous and visceral
adiposetissue.Inconjunction,markedalterationsofmetabolicgeneexpression
were shown consistent with upregulation of oxidative metabolism. This study
argues against chronic hypoxia being associated with adipose tissue
inflammation,andsimultaneouslysuggestsasignificantmetabolicshifttowards
increased fat oxidative metabolism during chronic hypoxia. Future studies are
warranted to investigate the effects of hypoxia on adipose tissue inflammation
andmetabolisminhumans.
UnhealthylifestyleofCOPDpatients:behindthesmokescreen
InChapter2,itisshownthatsmokingisassociatedwithdecreasedmuscle,fat
andbonemassesindependentlyoflungfunction.Thereby,smokingaloneposes
individuals at risk for accelerated physical deterioration. Smoking cessation is
oneofthehallmarksinCOPDtreatment,butitmaynotbesufficientintermsof
lifestyle management in COPD. It is increasingly becoming clear that COPD
patientsengageinanoverallpoorlifestylethatisalsocharacterizedbyphysical
159
SUMMARY
inactivityandanunhealthydiet.Importantly,thisconcernsnotonlythosewith
severeCOPDbutalsothosewithmilderdegreesofairflowobstruction.Chapter
3 demonstrates a 40% decreased daily physical activity in mild‐to‐moderate
COPD patients which is further characterized by fewer, shorter, and less
intensive bouts of moderate‐to‐vigorous exercise compared with those of
healthysubjects.Nearly70%oftheirtimeduringtheday,COPDpatientsspend
sedentary. Furthermore, Chapter 5 shows that only 8% of COPD patients
perform high‐intensity physical activities for more than 90 minutes per week.
Physical inactivity can lead to muscular deconditioning, adverse body fat
distribution and low‐gradesystemic inflammation, and may thereby contribute
todiseaseburdenandprogressionofCOPD.
The studied COPD patients in Chapter 5 had a poor dietary quality
characterizedbyincreasedfatconsumption,anddecreasedconsumptionoffruit,
vegetables,anddietaryfiber.Asystematicreviewoftheliterature(Chapter7)
further revealed that a greater intake of dietary fiber has been consistently
associated with better lung function in the general population, lower odds of
COPD,andlowerriskofincidentrespiratorysymptomsandCOPD.Importantly,
these associations were independent of smoking. The mechanisms by which
dietaryfiberleadstothesebeneficialeffectsrequirefurtherstudy.
EarlymetabolicriskinCOPD
Chapter8putsintoperspectivetheworkdescribedinthisthesisanddiscusses
future directions. Early loss of peripheral muscle oxidative phenotype and
adversebodyfatdistributionputthepatientwithCOPDatriskformetabolicand
cardiovascular diseases– already in those with mild pulmonary deficits. Future
research should elucidate whether this early metabolic risk is caused by COPD
itselforbyanunhealthylifestyle,andwhetherthetwoactsynergistically.
Animprovedunderstandingofthemechanismsofincreasedearlymetabolic
riskinpatientswithCOPDisprojectedtoresultinnewtherapeuticdirectionsto
reducediseaseburdenandmortality.
160
Samenvatting
161
SAMENVATTING
Wereldwijdlijden64miljoenmensenaandechronischelongziekteCOPD,ende
verwachting is dat COPD in het jaar 2020 doodsoorzaak nummer drie zal zijn.
COPDiseenverzamelnaamvooremfyseemenchronischebronchitis.Patiënten
met vergevorderd COPD overlijden hoofdzakelijk aan de gevolgen van het
primaire longlijden terwijl patiënten met mild‐tot‐matig COPD juist vaker
overlijdenaandegevolgenvanhart‐envaatziekten.Deverhoogdeziektelasten
sterfte bij COPD patiënten wordt waarschijnlijk bepaald door zowel ziekte‐
specifiekefactorenalsdooralgemenetrendsindebevolkingzoalsvergrijzingen
vervetting. Dit proefschrift richt zich op patiënten met mild‐tot‐matig COPD en
onderzoekt of er bij hen sprake is van een verhoogd metabool risico. Tevens
wordt onderzocht of veroudering en een ongezonde leefstijl hierin een rol
spelen.
Beenspierenalvroegopgerooktenuitgeblust
Verlies van spiermassa tijdens veroudering (genaamd ‘sarcopenie’) is normaal
en gaat gepaard met een achteruitgang van het lichamelijk functioneren. Veel
patiënten met COPD ervaren spierzwakte waardoor ze beperkt zijn in het
uitvoeren van dagelijkse activiteiten. De vraag rijst of deze spierzwakte
veroorzaakt wordt door de COPD of dat het ‘simpelweg’ te verklaren is door
sarcopenie. Hoofdstuk 2 toont aan dat de snelheid van het sarcopenieproces
vanaf het 70ste levensjaar gelijk is in patiënten met COPD en mensen zonder
COPD.Echter,voordietijdhebbenzereedsveelminderspiermassadangezonde
ouderen.Dushetverliesvanspiermassabegintaanzienlijkvroegerinhetleven
vandezepatiënten,ofwelzehaddenaltíjdalminderspiermassa.Hethoofdstuk
laattevensziendatouderendierokenmaargeenCOPDhebben,evenmagerzijn
als hun leeftijdsgenoten mét COPD. Als primaire risicofactor voor COPD heeft
rokendusnietalleeneennegatiefeffectopdelongenmaarookopspiermassa.
De mechanismen die leiden tot spierdisfunctie zijn goed bestudeerd in
patiënten met ernstig COPD, maar niet in patiënten met mildere vormen van
COPD. Hoofdstuk 3 toont bewijs voor een verminderd uithoudingsvermogen
van de bovenbeenspier in patiënten met mild‐tot‐matig COPD die nog een
normale spiermassa hebben. Het uithoudingsvermogen van de bovenbeenspier
was gerelateerd aan nadelige veranderingen in de stofwisseling van diezelfde
spier. Deze werden onder andere gekarakteriseerd door een verlies van het
aantal zuurstofafhankelijke spiervezels en een verlaagde genexpressie van
peroxisome proliferator‐activated receptor‐γ co‐activator 1α (een belangrijke
regulatorvanzuurstofafhankelijkestofwisseling).Dus,reedsinmilderevormen
van COPD bestaan er nadelige veranderingen van de stofwisseling van de
bovenbeenspiereneneenverminderduithoudingsvermogenvandiespieren.
162
SAMENVATTING
Vetalsontstekingsbron
In de afgelopen 10 jaar heeft wetenschappelijk onderzoek aangetoond dat
vetweefsel niet enkel een opslagplaats is voor energie, maar tegelijkertijd
eiwitten uitscheidt met ontstekingsactiviteit (genaamd ‘adipokines’). Er wordt
gesuggereerd dat een disregulatie van adipokine‐uitscheiding een rol speelt in
ontstekingsactiviteitbijCOPD.Hoofdstuk4presenteerteenvergelijkingvande
mate van vetweefselontsteking tussen patiënten met mild‐tot‐matig COPD en
gezondemensen.Inbioptenvanonderhuidsbuikvetbleekdatnochgenexpressie
van adipokines, noch het aantal ontstekingscellen (macrofagen) verschillend
waren tussen patiënten met de mild‐tot‐matig COPD en de gezonde mensen.
Evenmin waren er verschillen in bloedconcentraties van adipokines tussen de
patiënten en de gezonde mensen. Een hogere vetmassa hing wel samen met
meer vetweefselontsteking in zowel de COPD patiënten als in de gezonde
mensen.Dezeresultatensuggererendatontstekingvanhetonderhuidsebuikvet
over het algemeen niet afwijkend is in patiënten met mild‐tot‐matig COPD. Als
onderdeelvandemetabole‐risicobepalingbleekdatdepatiëntenmetCOPDwel
een verminderde insuline gevoeligheid hadden in vergelijking met de gezonde
mensen. Dit staat dus mogelijk los van de ontstekingsactiviteit in onderhuids
buikvet.
Hoofdstuk 5 laat zien dat patiënten met mild COPD een verhoogde
hoeveelheid vetweefsel ín hun buik hebben, onafhankelijk van obesitas of een
toegenomen buikomvang. Bloedconcentratie van het ontstekingseiwit
interleukine(IL)‐6waseveneensverhoogdenwaspositiefgeassocieerdmetde
hoeveelheid vetweefsel in de buik. Bovendien was de IL‐6 concentratie
voorspellend voor de hogere sterfte van de COPD patiënten. Deze resultaten
suggererendatinterventiesgerichtophetverminderenvanvetmassaíndebuik
een gunstig effect in COPD kunnen hebben door verlaging van systemische
ontstekingengeassocieerdemortaliteit.
Experimentele studies suggereren dat een laag zuurstofgehalte (hypoxie)
eenlokaleontstekingsreactieuitloktinvetweefselvanobesemuizen.Alshypoxie
inderdaad een van de mechanismen is die vetweefselontsteking kan
veroorzaken, kan dit relevant zijn voor patiënten met COPD aangezien zij vaak
hypoxie hebben door hun longfalen. In hoofdstuk 6 is geëxploreerd of muizen
dieblootgesteldwordenaanchronischehypoxiegekarakteriseerdwordendoor
vetweefselontsteking. Er werd echter een verlaagde en niet een verhoogde
vetweefselontstekinggevondeninzowelonderhuidsvetalsookbuikvetvandeze
muizen.Tevenswaserinhetvetweefseleenverhoogdeexpressievangenendie
belangrijk zijn voor de zuurstofwisseling. Deze studie pleit ertegen dat
chronische hypoxie gepaard gaat met vetweefselontsteking, en suggereert een
belangrijke verschuiving naar meer vetverbranding tijdens chronische hypoxie.
Toekomstige studies zijn nodig om te onderzoeken of deze resultaten ook
bevestigdwordenbijCOPDpatiëntenmethypoxie.
163
SAMENVATTING
OngezondeleefstijlinCOPD:achterhetrookgordijn
Inhoofdstuk2blijktdatrokengeassocieerdismetverminderdespier‐,vet‐en
botmassa,onafhankelijkvandelongfunctie.Daarmeeverhoogtrokenhetrisico
vooreenversneldeachteruitgangvanfysiekfunctioneren.Stoppenmetrokenis
een van de hoekstenen in de behandeling van COPD, maar het is mogelijk niet
genoeg in het kader van leefstijl optimalisatie. Het wordt in toenemende mate
bekend dat patiënten met COPD een algeheel ongezonde leefstijl hebben die
naastrokenookgekenmerktwordtdoorverminderdelichamelijkeactiviteiten
een ongezond voedingspatroon. Het is belangrijk te benadrukken dat dit niet
enkel patiënten met ernstig COPD betreft, maar ook al patiënten met mild‐tot‐
matigCOPD.Hoofdstuk3toonteenafnamevandagelijksefysiekeactiviteitvan
40%inpatiëntenmetmild‐tot‐matigCOPDinvergelijkingmetgezondemensen.
Bovendien hadden deze patiënten slechts weinig aaneengesloten perioden van
lichamelijke activiteit van een hoog genoeg niveau dat gezondheidswinst
oplevert zoals wandelen of sporten. Sterker nog: bijna 70% van de dag
spendeerdendepatiëntenzittend.Hoofdstuk5laatbovendienziendatslechts
8% van de patiënten met mild COPD meer dan 90 minuten per week hoog‐
intense fysieke activiteit beoefenen, tegenover 18% van mensen zonder COPD.
Lichamelijke inactiviteit leidt tot een deconditionering van spieren, een
ongunstigeverdelingvanlichaamsvetensystemischeontsteking,enkanviadie
wegenbijdragenaanziektelastenprogressieinCOPD.
In vergelijking met mensen zonder COPD, hadden de COPD patiënten
bestudeerd in Hoofdstuk 5 een slechte voedingskwaliteit welke werd
gekarakteriseerddooreenhogevetconsumptie,enverminderdeconsumptievan
fruit, groenten en vezels. Een systematisch overzicht van de literatuur
(Hoofdstuk 7) laat zien dat juist een hogere inname van vezels gerelateerd is
aan betere longfunctie in de algemene populatie, en een lagere kans op het
ontwikkelenvansymptomenvandeluchtwegenenhetontstaanvanCOPD.Deze
associaties staan los van de invloeden van roken. De mechanismen hoe
voedingsvezels tot deze gunstige effecten kunnen leiden, dienen verder
onderzochtteworden.
VroegmetaboolrisicoinCOPD
Hoofdstuk 8 zet de resultaten van dit proefschrift in perspectief en
bediscussieert toekomstige onderzoeksrichtingen. Nadelige verandering in de
stofwisseling van de beenspieren en een ongunstige verdeling van lichaamsvet
verhogen het risico op metabole en hart‐ en vaatcomplicaties in patiënten met
COPD– reeds in de vroege fase van hun ziekte. Toekomstig onderzoek zal
uitwijzenofditvroegmetaboolrisicoveroorzaaktwordtdoorCOPDzelf,ofdat
een algeheel ongezonde leefstijl hieraan ten grondslag ligt, en of COPD en een
ongezondeleefstijlelkaarversterken.
164
SAMENVATTING
Meer kennis over de mechanismen van verhoogd vroeg metabool risico in
patiëntenmetCOPDzalleidentotnieuwetherapeutischeoptiesomdeziektelast
ensterftevanCOPDteverminderen.
165
Dankwoord
167
DANKWOORD
Hetuitvoerenvanpromotieonderzoekisallesbehalveeenonemanshow.Opdeze
plaatswilikiedereenbedankendieheeftbijgedragenaandetotstandkomingvan
ditproefschrift.
Allereerstbenikallepatiëntenengezondevrijwilligerszeerdankbaarvoorhun
participatieinonsonderzoek.
Dan mijn promotor,prof. Annemie Schols. Het isal door zo velen gezegd, en ik
kanmedaaralleenmaarbijaansluiten;jouwwetenschappelijkenthousiasmeis
benijdenswaardig en enorm inspirerend. Je gaf me de mogelijkheid om in
Amerika onderzoek te doen, waarvoor ik je zeer dankbaar ben. Mede door de
enorm hoge snelheid waarmee je mijn e‐mails beantwoordde, heeft mijn
onderzoek in een constante stroomversnelling gestaan. Soms leek het wel of je
met Harry Gosker, mijn co‐promotor, een wedstrijdje deed wie het snelst
reageerde. Harry, ontzettend bedankt voor al je inhoudelijke en praktische
support.Ikkonaltijdbijjebinnenlopenengeenwetenschappelijkeuitdagingis
jetegek.AnnemieenHarry,ikwiljulliehartelijkbedankenvoordeleerzameen
fijnejaren.Datweopeenzaterdagtotdiepindenachtbrainstormdenoverde
stellingen was tekenend voor jullie toewijding en betrokkenheid. Ik ben zeer
verheugd dat we een subsidie van het Longfonds hebben binnengehaald zodat
weonzesamenwerkingkunnenvervolgen.
Ik wil de leden van de beoordelingscommissie bedanken voor hun kritische
beoordelingengoedkeuringvanditproefschrift.
Prof.EmielWouters,uintroduceerdemijinditonderzoeksprojectwaarvooriku
dankbaarben.
Top Institute Pharma en alle partners binnen ons project (Maastricht UMC+,
UMC Utrecht, UMC Groningen, AstraZeneca, GlaxoSmithKline, Takeda, en
Danone) ben ik dankbaar voor de vruchtbare samenwerking en leerzame
bijeenkomsten.
ValéryHellwig,roomie,watmoestikzonderjou?Waarligtdiemap?Watmoetik
met dit formulier? Jouw logistieke en praktische bijdragen aan de klinische
studie zijn zeer waardevol geweest. We begonnen vrijwel tegelijkertijd op dit
project en hebben elkaar de weg gewezen in onderzoeksland. Het lag voor de
handdatjijparanimfwerd,dankjewelvooralles!
Paul Popelier, een vanzelfsprekendheid dat jij paranimf werd. Bedankt voor je
hulpmetPrezi.
168
DANKWOORD
Aging and for introducing me to the Health ABC Study. Annemarie Koster,
bedanktvoordefijnetijd,voorjeintensievebegeleidingenvoorjehulpbijalle
bureaucratierondomhetverwezenlijkenvaneentijdelijkebaaninAmerika.
ThankstoalltheHealthABCStudyco‐authorsforyourvaluableinput.
RamonLangen,bedanktvoordiekeerdatjemijnsamplesuitdekapottevriezer
hebt gered! Tevens bedankt voor de vele brainstormsessies. En laten we de
volgendekeermaarpassenvoordieCatalaansestierenballen.
Bettine Vosse, bedankt voor het opstarten van de klinische studie. Zowel in de
aanvraag voor de medisch ethische commissie als in de praktische
voorbereidingenhebjijveelwerkverricht.
Chiel de Theije, jij voerde de muisstudies uit voor je eigen proefschrift.
Regelmatighebiknaastjegezetenomdeweefselsteverzamelen.Hetlagvoorde
hand om een samenwerking aan te gaan om het vetweefsel van deze muizen
onderdeloeptenemen,wateenmooihoofdstukheeftopgeleverd.Bedankt.
IlseSlot,bedanktvoorjeinspanningendiekwalitatiefhoogstaandelaboratorium
analysesopleverden.
Frank Snepvangers en Marco Kelders, bedankt voor jullie expertise en het
verlenenvanassistentiebijdeverschillendelaboratoriumanalyses.
MijndankgaatuitnaardelongartsenvanhetMaastrichtUMC+,inhetbijzonder
prof. Geertjan Wesseling, voor de rekrutering van COPD patiënten voor de
klinischestudie.
GijsGoossens,bedanktvoorjeinzichtendierichtinggavenaandemuisstudie.
Juanita Vernooy, bij jou deed ik een vrijwillige stage en mijn afstudeerstage in
het lab van de Longziekten en raakte ik besmet met het onderzoeksvirus.
Bedanktdatjeeen‘goedwoordje’voormedeed.
Frits Franssen, bedankt voor het advies om eerst promotie onderzoek te doen
vóór de klinische opleiding. Ik zou dit op mijn beurt ook iedereen willen
aanraden.
Agnes Boots, bedankt voor je inspanningen met de Luminex analyses en je
immersnellereacties.
169
DANKWOORD
Een aantal publicaties is uiteindelijk niet in mijn proefschrift terechtgekomen.
Desalniettemin wil ik prof. Maurice Zeegers, Nicole Souren en de overige co‐
auteursbedankenvoorhunbetrokkenheidhierbij.
Bedanktaanallecollega’svandeafdelingLongziektenvoordeprettigejaren.
Beste Tijs, lieve broer, bedankt voor alle tijd en energie die je stak in het
ontwerpenvandecovervanmijnproefschrift.Hetisprachtiggeworden!
LieveBertyenMartin,bedanktvoorjulliesupporteninteresseindeafgelopen
jaren.
Tot slot, lieve Sharon, het is een hele tocht geweest maar nu is het
promotieonderzoekdanéchtklaar.Ikwiljeenormbedankenvoorallesteunin
deze jaren en voor de vrijheid die je me gunde om een tijdje in Amerika te
wonen. Hopelijk breekt er nu een rustigere periode aan met onze mooie zoon
Nathan.
170
CurriculumVitae
171
CURRICULUMVITAE
BramvandenBorst(18juli1983)behaaldein2001zijnVWOdiplomaaanhet
Van Maerlant Lyceum in Eindhoven. In 2002 kon hij starten met de opleiding
Geneeskunde aan de Universiteit Maastricht. Tijdens zijn opleiding deed hij
stages aan de Universiteit van Linköping in Zweden en de Royal Victoria
Infirmary, Newcastle in Engeland. Zijn afstudeerjaar stond geheel in het teken
vanlongziekten,enin2008behaaldehijzijnartsendiploma.Van2008tot2012
werkte hij als promovendus binnen de onderzoeksschool NUTRIM bij de
vakgroep Longziekten van het Maastricht University Medical Center+ onder
supervisievanprof.dr.ir.AnnemieScholsendr.HarryGosker.Eendeelvanhet
onderzoek dat hij in die periode uitvoerde staat beschreven in dit proefschrift.
Twee hoofdstukken komen voort uit onderzoek dat hij deed aan het National
Institute on Aging in Bethesda, USA. Sinds juli 2012 volgt hij de
specialistenopleidingtotlongartsinhetOrbisMedischCentruminSittard‐Geleen
enhetMaastrichtUniversityMedicalCenter+.Voortvloeienduitditproefschrift,
en financieel gesteund door het Longfonds, zal in 2013 een vervolgonderzoek
van start gaan met als titel “Vermindering van het verhoogde metabole en
cardiovasculairerisicobijCOPDpatiëntenmetovergewicht.”
173
Listofpublications
175
LISTOFPUBLICATIONS
Originalresearcharticlesandreviews
BramvandenBorst,AnnemieMWJSchols,ChieldeTheije,SEleonoreKoehler,
Agnes W Boots, Gijs H Goossens, Harry R Gosker. Characterization of the
inflammatory and metabolic profile of adipose tissue in a mouse model of
chronichypoxia.JournalofAppliedPhysiology2013Mar28[Epubaheadofprint],
doi:jap.00460.2012
Bram van den Borst, Harry R Gosker, Annemie MWJ Schols. Central fat and
peripheralmuscle:partnersincrimeinchronicobstructivepulmonarydisease.
AmericanJournalofRespiratoryandCriticalCareMedicine2013;187(1):8‐13
BramvandenBorst,IlseGMSlot,ValéryACVHellwig,BettineAHVosse,Marco
CJM Kelders, Esther Barreiro, Annemie MWJ Schols, Harry R Gosker. Loss of
quadriceps muscle oxidative phenotype and decreased endurance in patients
with mild‐to‐moderate COPD. Journal of Applied Physiology 2012 Jul 19 [Epub
aheadofprint],doi:jap.00508.2012
BramvandenBorst,HarryRGosker,AnnemarieKoster,BinbingYu,StephenB
Kritchevsky, Yongmei Liu, Bernd Meibohm, Thomas B Rice, Michael Shlipak,
Sachin Yende, Tamara B Harris, Annemie MWJ Schols. The influence of
abdominal visceral fat on inflammatory pathways and mortality risk in
obstructivelungdisease.AmericanJournalofClinicalNutrition2012;96(3):516‐26
BramvandenBorst,HarryRGosker,GeertjanWesseling,WilcodeJager,Valéry
ACV Hellwig, Frank J Snepvangers, Annemie MWJ Schols. Low‐grade adipose
tissue inflammation in patients with mild‐to‐moderate chronic obstructive
pulmonarydisease.AmericanJournalofClinicalNutrition2011;94(6):1504‐12
Bram van den Borst, Nicole YP Souren, Ruth JF Loos, Aimée DC Paulussen,
Catherine Derom, Annemie MWJ Schols, Maurice P Zeegers. Genetics of
maximally attained lung function: a role for leptin? Respiratory Medicine
2012;106(2):235‐42
Bram van den Borst, Annemarie Koster, Binbing Yu, Harry R Gosker, Bernd
Meibohm, Douglas C Bauer, Stephen B Kritchevsky, Yongmei Liu, Anne B
Newman, Tamara B Harris, Annemie MWJ Schols. Is age‐related decline in lean
massandphysicalfunctionacceleratedbyobstructivelungdiseaseorsmoking?
Thorax2011;66(11):961‐9
BramvandenBorst,HarryRGosker,MauricePZeegers,AnnemieMWJSchols.
Pulmonaryfunctionindiabetes:ametaanalysis.Chest2010;138(2):393‐406
176
LISTOFPUBLICATIONS
Eric LA Fonseca Wald, Bram van den Borst, Harry R Gosker, Annemie MWJ
Schols.DietaryfiberandfattyacidsinCOPDriskandprogression:asystematic
review.[submitted]
BramvandenBorst,IlseGMSlot,ValéryACVHellwig,EstherBarreiro,Annemie
MWJSchols,HarryRGosker.Normalmuscleoxidativemetabolicgeneresponse
toO2‐desaturatingexerciseinCOPD.[submitted]
Casereport
BramvandenBorst,MichieldeVries,WieldeLange.Amanwithatumoronhis
back[inDutch].TijdschriftvoorGeneeskunde2008;64(18):932‐4
Peer‐reviewedletterstotheeditor
BramvandenBorst,GeertjanWesseling.AcuteexacerbationsinCOPD:it’sthe
weekend but it can’t wait until Monday. European Respiratory Journal
2012;39(6):1547
Bram van den Borst, Annemie MWJ Schols. Low bone mineral density in
emphysema: epiphenomenon of a wasting phenotype? American Journal of
RespiratoryandCriticalCareMedicine2011;184(9):1087‐8
Bram van den Borst, Nicole YP Souren, Marij Gielen, Ruth JF Loos, Aimée DC
Paulussen, Catherine Derom, Annemie MWJ Schols, Maurice P Zeegers.
AssociationbetweentheIL6‐174G/CSNPandmaximallyattainedlungfunction.
Thorax2011;66(2):179
Bookchapter
Bram van den Borst, Annemie MWJ Schols. Chronic Obstructive Pulmonary
Disease. Co‐Morbidities and Systemic Consequences. 1st ed. New York (NY):
HumanaPress;2012.Chapter11,BodycompositionabnormalitiesinCOPD;Pp.
171‐92
177

Early metabolic risk in COPD

Transcription

Similar documents

Darrell Green - American Thoracic Society

An Everyday Guide for You and Your Family

feed the snack attack! - Coventry Health Care of Missouri

TLS® Thermochrome with Advantra Z® and

YOGURT ACTIVIA LIGHT VANILLA LOW FAT Nutrition

Gut-Brain Regeneration

Top 10 Tips: Improve everyday nutrition

Techniques and Principles

Resolution help from the experts

2016 Blue Care Network - UAW Retiree Medical Benefits Trust